JP6924104B2 - Electricity price forecasting system - Google Patents

Description

The present disclosure relates to an electric power price forecasting system that predicts the contracted price of electric power traded in the electric power trading market.

With the full-scale expansion of electricity liberalization in recent years, it is expected that the focus of electricity transactions will shift from bidding transactions to bidding transactions. Electric power bid transactions are conducted on the electric power exchange. For example, on the Japan Electric Power Exchange (JEPX), electric power transactions on any future delivery date (1 day) are performed on the day before market (spot market) and on the day market. Through multiple types of markets, such as the pre-hours market and the forward market, where electricity is traded before the day before, every metering unit of electricity (0-30 minutes per hour, 30-60 minutes). It is said. More specifically, electricity on any delivery date will be traded in the forward market at a larger particle size, such as one month or one week, for up to three years, one day before the delivery date. Final adjustment in the market. However, the power plan that was finalized one day ago cannot be avoided due to, for example, a stoppage due to an accident at a power plant (power plant) or an increase in power demand due to unexpected changes in temperature. Adjustments are made after the day-to-day plan is formulated.

Further, in a bidding transaction, a bid is executed when the bidding conditions (price, quantity, etc.) of a power selling business such as a power generation business and a power purchasing business such as a retail business are matched. For example, in the forward market, it is possible to execute a fixed price on a monthly or weekly basis. On the other hand, in the market one day before, after the bidding deadline, bids are divided into buying and selling for each electric product, which is the measurement unit (48 hours) of electric power for one day, and the supply and demand curve (amount of selling-price). The contract price and the contracted power amount are determined for each power product at the intersection of the line and the amount of purchase-price line). On the other hand, in the market on the day, it is a rough trading for each electric power product, and it is traded not only by limit orders that specify the price under the principle of price priority and time zone priority, but also by market orders that do not specify the price. NS. In other words, the contract price in the one-day market and the same-day market is determined according to the bidding status for each electric power product.

Conventionally, a method of predicting a contract price in such an electric power trading market by a computer has been proposed (for example, Patent Documents 1 to 4).
In Patent Document 1, a regression analysis is performed on past weather record data, past transmission reserve capacity record data of an interconnection line connecting a plurality of supply areas, and past transaction price record data, and future weather forecast data and the above-mentioned plurality. It is disclosed that a regression equation is obtained using the future transmission reserve capacity prediction data of the interconnection line connecting the supply areas of the above as an explanatory variable and the above predicted market price as an objective variable, and the predicted market price is obtained based on the regression equation. Further, in Patent Document 1, when creating bid data, in-house demand data including the obtained forecast market price, in-house power generation equipment data, and electric power forward contract data is acquired, and the amount (electric energy) in the bid pair in each bid block. Amount) is set.

In Patent Document 2, the prediction unit that calculates the predicted contract price and the predicted contract quantification under a plurality of contract prediction conditions according to the factor information, which is the information of the factors that influence the contract of the electric power transaction, has previously calculated the predicted contract quantification. Calculate the predicted contract price based on, and calculate the predicted contract price and the predicted contracted amount as statistical information based on the contracted actual price, the contracted amount, and the contracted actual information accumulated factor information in the past contracts. It is stated that. Further, Patent Document 2 also describes that different weather conditions such as sunny weather, cloudy weather, and rainy weather are defined as contract prediction conditions.

Patent Document 3 includes a power generation cost calculation means for calculating the power generation cost calculated by creating a power supply and demand plan based on the demand estimation data and the power supply data of the power market participants, and the power generation cost calculation means. A bid planning means that determines the bid amount and bid price for the spot market based on the power generation cost and creates a bid plan, and a consignable amount of the interconnection line based on the bid amount and bid price determined by the bid planning means. It is equipped with a spot market contracting means for determining transaction volume and transaction price using data on, and a bilateral / forward contract setting means for setting bilateral contracts and forward contracts, and is equipped with a power generation cost calculation means, a bid planning means, and a spot market. It is disclosed that the contracting means performs processing that reflects the contract set by the bilateral / forward contract setting means. More specifically, it is described that the respective capacities of the bilateral contract and the forward contract are added to the demand estimation data.

Japanese Unexamined Patent Publication No. 2006-172246 Japanese Unexamined Patent Publication No. 2016-164712 JP-A-2007-65954

In the free electricity trading market (for example, JEPX), electricity sellers usually want to sell at the highest possible price while considering fluctuations in power generation capacity due to equipment failures and avoidance of supply shortages that impose penalties. In addition, the power purchaser decides the bid conditions with the intention of buying at the lowest possible price while considering fluctuations in electricity demand due to climate change and avoidance of supply shortages. However, in markets such as the one-day-ahead market and the same-day market where the contracted price of electricity fluctuates according to the bidding situation (hereinafter referred to as the fluctuating market), businesses that do not have the know-how related to bidding transactions should set particularly appropriate bidding conditions. difficult. As a result, there is a risk that business operations will become difficult due to increased costs and reduced profits, which hinders the activation of free electricity transactions.

In view of the above circumstances, at least one embodiment of the present invention aims to provide an electric power price forecasting system that predicts a contract price in a fluctuating market such as a one-day market or a same-day market.

(1) The electric power price prediction system according to at least one embodiment of the present invention is
A power price prediction system that calculates the contract forecast price in a variable market where the market price of electricity delivered from the power seller to the power purchaser fluctuates depending on the bidding status on the target delivery date and time, which is the desired delivery date and time. There,
Environmental performance information including past weather information, about quantitative performance information that is the sum of past power traded in each of the variable market and the forward market that trades power before the trading date of the variable market, said. A past record database that stores the contract price record information, which is the record of the contract price of electric power traded in the past in the variable market, by the delivery date and time and by the bid area, and
A forward contract total amount acquisition unit that acquires the forward contract total amount, which is the amount of electric power delivered in the target bid area, which is the desired bid area at the target delivery date and time traded in the forward market.
An environmental forecast acquisition unit that acquires environmental forecast information including weather information of the target delivery date and time in the target bid area, and
Based on the comparison between the environmental forecast information of the target delivery date and time in the target bid area and the environmental record information stored in the past record database, the target delivery date and time in the target bid area can be obtained from the past record database. The Approximate Quantitative Results Acquisition Department that acquires the Approximate Quantitative Results Information corresponding to the time,
By subtracting the total amount of forward contracts from the actual quantitative information, the quantitative forecast calculation unit that calculates the quantitative forecast, which is the predicted value of the electric energy traded in the variable market,
It is provided with a contract forecast price calculation unit that acquires the contract price actual information from the past record database based on the contract quantitative forecast and calculates the contract price forecast price based on the acquired contract price actual information.

In variable markets such as the one-day market (spot market) and the same-day market (hourly market) where electricity is traded on the delivery date, the delivery date is divided into time zones such as every 30 minutes (total of 48 time zones). ), The power purchaser and the power purchaser each place a bid or sell bid by specifying the price and quantity for each time zone, but the conditions of these bids and the conditions of the sell bid must match. Since the contract is made with, the market price of electricity (contract price) fluctuates according to the bidding situation.

According to the configuration of (1) above, the electric power price prediction system bases the contract forecast price of the liquid market at the desired delivery date and time (target delivery date and time) to be predicted based on the past record information stored in the past record database. Predict. Specifically, the approximate quantitative actual information associated with the environmental actual information that matches (same or similar) the environmental forecast information of the target delivery date and time is used as the predicted electric energy to be delivered on the target delivery date and time. In addition, by subtracting the amount of electric power traded in the forward market (total amount of forward contracts) from the predicted electric energy with respect to the target delivery date and time, the amount of electric power traded (contracted) in the variable market is predicted.

Then, for example, if the past contract price actual information stored in the past actual database is associated with the result of the approximately quantitative forecast calculated in the past (approx. Quantitative forecast actual) and stored, the approximately quantitative forecast is performed. The contract forecast price of the target delivery date and time is calculated based on the contract price actual information, such as acquiring the contract price actual information associated with the contract price forecast result that matches the above and using it as the contract price of the target delivery date and time. In this way, it is possible to accurately predict the contract forecast price at the desired delivery date and time based on the past results. In addition, by making it possible to calculate the contract forecast price with high accuracy, it is possible to present the optimum buy / sell bid in the variable market.

(2) In some embodiments, in the configuration of (1) above,
The past performance database further stores the contract price performance information in association with the contracted price forecast calculation unit, which is the contracted quantitative forecast calculated in the past, or the contracted price forecast information. And
The contract price forecast calculation unit obtains the contract price actual information from the past actual database based on the comparison between the contract quantitative forecast and the contract quantitative forecast actual or the contract quantitative actual information stored in the past actual database. get.
According to the configuration of (2) above, in the past record database, the contract price actual information at the past delivery date and time is used as the calculation result (approx. It is stored in association with each other, and the contract forecast price calculation unit acquires the contract price actual information from the past actual database based on the comparison between the contract quantitative forecast and the contract quantitative forecast actual or the contract quantitative actual information. As a result, the contract price actual information used for calculating the contract forecast price can be appropriately acquired from the past record database, and the contract forecast price can be calculated.

(3) In some embodiments, in the configuration of (2) above,
The fuel price acquisition department that acquires fuel price information of the fuel used for power generation by the power seller,
Further equipped with a power generation capacity prediction acquisition unit for acquiring a power generation capacity prediction at the target delivery date and time of the power plant owned by the power sales business operator.
The past record database is composed of the environmental record information, the contracted quantitative record information, the contract price record information, the power generation capacity record information for each past date and time in the power seller, and the fuel price information in the past. Along with memorizing a certain fuel price actual information by associating it with each other
The said about quantitative result acquisition department
Based on the comparison between the environmental forecast information and the environmental performance information, the buy bid contracted quantitative result acquisition unit that acquires the contracted quantitative result information from the past record database and uses the bid bid contracted quantitative result as the buy bid contracted quantitative result.
In addition to the comparison between the environmental forecast information and the environmental actual information, the past is based on the comparison between the power generation capacity forecast and the power generation capacity actual information and the comparison between the fuel price information and the fuel price actual information. It has a selling bid contract quantitative result acquisition unit that acquires the above-mentioned approximate quantitative record information from the record database and uses it as a sell bid contract quantitative record.
The approximate quantitative prediction calculation unit
The Bid Bid Quantitative Prediction Calculator, which calculates the Bid Quantitative Prediction by subtracting the total amount of the forward contracts from the Bid Quantitative Results.
It has a sell bid contract quantitative forecast calculation unit that calculates a sell bid contract quantitative forecast by subtracting the forward contract total amount from the sell bid contract quantitative performance information.

According to the configuration of (3) above, the Approximate Quantitative Results Acquisition Department calculates the Approximate Quantitative Results Information on Buy Bid and Sell Bid, respectively, and the Approximate Quantitative Prediction Calculation Department buys the Approximate Quantitative Prediction (Buy Bid Approximate Quantitative Prediction). And sell bid (sell bid contract quantitative forecast) are calculated separately. Then, the buy bid contract quantitative forecast and the sell bid contract quantitative forecast are calculated by subtracting the total amount of the forward contracts from the contract quantitative performance information for each of the buy bid and the sell bid. In addition, the past record database corresponds to the bid bid contract quantitative forecast (buy bid contract quantitative forecast result), the sell bid contract quantitative forecast (sell bid contract quantitative forecast result), and the contract price actual information calculated in the past. The contract price is calculated based on the contract price performance information when the bid bid contract quantitative forecast result and the sell bid contract quantitative forecast result match the bid bid contract quantitative forecast and the sell bid contract quantitative forecast, respectively. calculate.

As described above, the purchase bid contract quantitative result is the contract quantitative record information when the environmental record information (past) matches the environmental forecast information of the target delivery date and time. In other words, it corresponds to the amount of power demand on the target delivery date and time predicted based on environmental factors. Therefore, the quantitative forecast of bids and bids corresponds to the forecast of electricity demand in a variable market. On the other hand, as for the sales bid contract quantitative results, each of the environmental record information (past), the power generation capacity record information (past), and the fuel price record information (past) are the environmental forecast information of the target delivery date and time, the power generation capacity forecast, and the power generation capacity forecast, respectively. It is about quantitative actual information when it matches with the fuel price information. In other words, it corresponds to the amount of power supplied on the delivery date predicted based on environmental factors and power generation circumstances factors. Therefore, the bid contract quantitative forecast Qcb corresponds to the forecast of the amount of power supply in the variable market. Therefore, by referring to the past performance database, the contract price actually executed for the power demand forecast (buy bid contract quantitative forecast result) and power supply amount forecast (sell bid contract quantitative forecast result) performed in the past. By obtaining the actual information, the accuracy of the contract forecast price calculated based on the contract price actual information can be further improved.

(4) In some embodiments, in the configuration of (3) above,
A power source configuration database that stores power source configuration information, which is information on equipment constituting the power plant owned by the power seller, and stop information of the power plant for each power seller, and a power source configuration database.
A power generation capacity calculation unit that calculates the power generation capacity of the power sales business operator based on the power generation configuration information and the stop information of the power plant is further provided.
The power generation capacity prediction acquisition unit acquires the power generation capacity calculated by the power generation capacity calculation unit as the power generation capacity prediction.
According to the configuration of (4) above, the power generation capacity prediction of the target delivery date and time in the target bidding area can be appropriately calculated based on the power generation configuration information and the stop information of the power plant. Therefore, it is possible to obtain the sales bid contract quantitative result more accurately.

(5) In some embodiments, in the configurations (3) to (4) above,
In the past, the past performance database used the buy bid contract quantitative forecast result, which is the bid bid contract quantitative forecast calculated based on the bid bid contract quantitative forecast, to calculate the bid bid contract quantitative forecast result. It is stored in association with the said quantitative result information, which is the purchased bid contract quantitative result.
Based on the comparison between the buy bid contract quantitative forecast and the buy bid contract quantitative forecast result stored in the past record database, the buy bid contract quantitative forecast result and the contract quantitative result information are obtained from the past record database. In addition to being acquired, it is further equipped with a buy bid contract quantitative prediction error calculation unit that calculates a buy bid contract quantitative forecast error based on the acquired bid bid contract quantitative forecast result and the contract quantitative result information.
The bid bid contract quantitative prediction calculation unit calculates the bid bid contract quantitative forecast using the bid bid contract quantitative forecast error.
According to the configuration of (5) above, the buy bid contract quantitative prediction is corrected by the bid bid contract quantitative prediction error. This makes it possible to improve the accuracy of the buy bid contract quantitative prediction.

(6) In some embodiments, in the configurations (3) to (5) above,
In the past, the past performance database uses the sell bid contract quantitative forecast result, which is the sell bid contract quantitative forecast calculated based on the sell bid contract quantitative forecast, to calculate the sell bid contract quantitative forecast result. It is stored in association with the above-mentioned about quantitative result information, which is the sell bid about quantitative result.
Based on the comparison between the sell bid contract quantitative forecast and the sell bid contract quantitative forecast result stored in the past record database, the sell bid contract quantitative forecast result and the contract quantitative result information are obtained from the past record database. In addition to being acquired, it is further equipped with a sell bid contract quantitative prediction error calculation unit that calculates a sell bid contract quantitative forecast error based on the acquired sell bid contract quantitative forecast result and the contract quantitative result information.
The sell bid contract quantitative prediction calculation unit calculates the sell bid contract quantitative forecast using the sell bid contract quantitative forecast error.
According to the configuration of (6) above, the sell bid contract quantitative prediction is corrected by the sell bid contract quantitative prediction error. As a result, the accuracy of the sell bid contract quantitative prediction can be improved.

(7) In some embodiments, in the configuration of (3) above,
A power transmission destination information database for storing power transmission destination information including the location of the power transmission destination related to the power transmission destination to which the power purchaser transmits the electric power, and a power transmission destination information database.
Further, a power transmission destination information acquisition unit for acquiring the power transmission destination information of a plurality of the power transmission destinations located in the target bidding area from the power transmission destination information database is further provided.
The environmental forecast acquisition unit acquires the environmental forecast information for each of the locations of the plurality of power transmission destinations, and obtains the environmental forecast information.
The purchase bid contract quantitative forecast calculation unit is based on the comparison between each of the environmental forecast information for each location of the transmission destination and the environmental record information stored in the past record database, and from the past record database. The total of the obtained quantitative results information is used as the buy bid contract quantitative forecast.
According to the configuration of (7) above, each power purchaser obtains about quantitative actual information for each power transmission destination (contractor) who has contracted for power supply, and the total is used as the buy bid contract quantitative actual. In this way, the calculation accuracy can be improved by accumulating the electric power demand for each contractor customer predicted based on the power transmission destination information and calculating the final bid bid contract quantitative prediction.

(8) In some embodiments, in the configurations (3) to (7) above,
The variable market includes the same-day market that determines the market price on the delivery date and the one-day market that determines the market price one day before the delivery date.
The past record database contains the environmental record information, the buy bid contract quantitative forecast record, the sell bid contract quantitative forecast record, and the current day contract quantitative record which is the electric energy for each past date and time traded in the market on the day. The information and the one-day-ahead quantitative actual information, which is the amount of electricity for each past date and time traded in the one-day-ahead market, are stored in association with each other.
The electricity price forecasting system further
Based on the comparison between the bid bid contract quantitative forecast and the environmental forecast information, and the bid bid contract quantitative forecast result and the environmental record information stored in the past record database, the same-day contract quantitative result information and the above-mentioned one. The Bid Breakdown Results Acquisition Department, which acquires information on the quantitative results of the day before,
Based on the current-day contract quantitative performance information and the one-day-previous quantitative performance information acquired by the bid breakdown performance acquisition unit, the above-mentioned traded in each of the same-day market and the one-day-ahead market constituting the variable market. In addition to calculating the purchase bid breakdown results, which is the breakdown of the amount of electricity, the buy bid market breakdown forecast, which is the breakdown of the same-day market and the one-day-ahead market in the purchase bid contract quantitative forecast, is calculated based on the purchase bid breakdown results. Bid market breakdown forecast calculation department and
Based on the comparison between the sold bid contract quantitative forecast and the environmental forecast information, and the sold bid contract quantitative forecast result and the environmental record information stored in the past record database, the same-day contract quantitative result information and the above-mentioned one. The variable market is constructed based on the sales bid breakdown performance acquisition unit that acquires the day-ahead quantitative performance information, the current-day contract quantitative performance information acquired by the sales bid breakdown performance acquisition department, and the one-day contract quantitative performance information. The current bid breakdown, which is a breakdown of the amount of electricity traded in each of the same-day market and the one-day-ahead market, is calculated, and the same-day market in the quantitative forecast of bid contracts is calculated based on the bid breakdown record information. And a bid bid market breakdown forecast calculation unit that calculates the bid bid market breakdown forecast, which is the breakdown of the previous day's market.
The contract forecast price calculation unit is based on the buy bid contract quantitative forecast, the bid market breakdown forecast, the sell bid contract quantitative forecast, and the sell bid market breakdown forecast, and at least one of the same-day market and the one-day-ahead market. The contract forecast price in the above is calculated.
According to the configuration of (8) above, the market price forecast can be calculated for each market such as the same-day market and the one-day-ahead market constituting the variable market.

(9) In some embodiments, in the configuration of (8) above,
The past performance database further indicates the past power consumption of the inter-regional interconnection line connecting the regions composed of one or more of the bidding areas with respect to the approximately quantitative performance information. The information is associated and memorized,
The interconnection line allowance acquisition unit that acquires the transmission allowance of the inter-regional interconnection line, and the interconnection line allowance acquisition unit.
Based on the comparison between the buy bid contract quantitative forecast and the sell bid contract quantitative forecast, and the buy bid contract quantitative forecast result and the sell bid contract quantitative forecast result stored in the past performance database, the past The interconnection line usage record acquisition unit that acquires the interconnection line usage record information from the track record database, and
A division prediction execution unit that predicts whether or not a market division will occur based on a comparison between the transmission capacity of the inter-regional interconnection line and the actual information on the usage amount of the interconnection line.
An area forecast price calculation unit for calculating an area forecast price, which is the contract forecast price for each bid area when the market division is predicted to occur, is further provided.
According to the configuration of (9) above, it is possible to calculate the contract forecast price for each bidding area when the market division occurs, and the contract forecast price can be calculated more accurately.

(10) In some embodiments, in the configuration of (9) above,
An output unit for outputting the position information of the inter-regional interconnection line from the result of the prediction by the division prediction execution unit is further provided.
According to the configuration of (10) above, the result of the market division prediction is displayed together with the position information of the interregional interconnection line, such as showing the result of the market division prediction in the map information showing the position of the interregional interconnection line. By outputting to an output device such as, it is possible to grasp the risk of market division at a glance.

(11) In some embodiments, in the configurations (1) to (10) above,
A bid pattern determination unit for determining a bid pattern for a plurality of markets constituting the variable market is further provided.
According to the configuration of (11) above, an appropriate bid can be placed in a variable market.

(12) In some embodiments, in the configurations (1) to (11) above,
A bid condition determination unit for determining bid conditions in consideration of the operation record information of the power plant including the power generation unit price record generated by the power plant owned by the power seller is further provided.
According to the configuration of (12) above, the bidding conditions that are optimal for the operation of the power plant can be determined by determining the bidding conditions in consideration of the operation record information of the power plant.

(13) In some embodiments, in the configuration of (9) above,
The interconnection line usage record acquisition unit includes the buy bid contract quantitative forecast and the sell bid contract quantitative forecast, and the buy bid contract quantitative forecast result and the sell bid contract quantitative forecast result stored in the past record database. The degree of similarity with is obtained for each record managed in the past record database, and the interconnection line usage record information is acquired from the past record database based on the similarity.

(14) In some embodiments, in the configuration of (8) above,
The purchase bid breakdown result acquisition unit determines the degree of similarity between the purchase bid contract quantitative forecast and the environmental forecast information and the buy bid contract quantitative forecast record and the environmental record information stored in the past record database. Obtained for each record managed in the performance database, and based on the similarity, the current day's quantitative performance information and the one day's previous quantitative performance information are acquired from the past performance database.
The selling bid breakdown result acquisition unit records the degree of similarity between the selling bid contract quantitative forecast and the environmental forecast information and the selling bid contract quantitative forecast result and the environmental performance information stored in the past record database. It is obtained for each time, and based on the similarity, the current day's quantitative actual information and the one day's previous quantitative actual information are acquired from the past record database.

(15) In some embodiments, in the configurations (2) to (14) above,
The contract forecast price calculation unit determines the degree of similarity between the contract forecast and the contract forecast result or the contract quantitative record information stored in the past record database for each record managed in the past record database. Obtain the contract price actual information from the past actual database based on the similarity.

(16) In some embodiments, in the configurations (1) to (15) above,
The approximately quantitative result acquisition unit determines the degree of similarity between the acquisition condition including the environmental forecast information of the target delivery date and time in the target bid area and the information including the environmental record information stored in the past record database. It is obtained for each record managed in the past record database, and based on the similarity, the approximate quantitative record information corresponding to the time of the target delivery date and time in the target bid area is acquired from the past record database.

According to the configurations (13) to (16) above, various information is acquired from various databases by using, for example, matching by dynamic programming or machine learning (deep machine learning). As a result, it is possible to accurately predict the contract forecast price at the desired delivery date and time based on the past results.

(17) In some embodiments, in the configuration of (14) above,
The buy bid breakdown record acquisition unit and the sell bid breakdown record acquisition unit obtain the same-day contracted quantitative record information and the one-day-a-day previous quantitative record information included in the record exceeding the preset threshold of similarity. Obtained from the past record database,
The buy / bid market breakdown forecast calculation unit and the sell / bid market breakdown forecast calculation unit are a set of a plurality of acquired daily quantitative actual information when a plurality of the records have the same degree of similarity exceeding the threshold value, respectively. This is a value calculated by performing a predetermined calculation for each of the same-day about quantitative result set and the one-day-a-day about quantitative result set, which is a set of a plurality of acquired one-day-around quantitative results information. The buy bid breakdown result or the sell bid breakdown result is calculated based on the current day's approximate quantitative representative value based on the set and the one-day's approximate quantitative result representative value based on the one-day advance quantitative result set.

(18) In some embodiments, in the configuration of (15), the contract forecast price calculation unit obtains the contract price actual information included in the record exceeding the preset threshold value of the similarity. Obtained from the past record database and
When a plurality of the records have the similarity exceeding the threshold value, a predetermined calculation is performed on the contract price actual set which is a set of the contract price actual information to be acquired, and the contract price actual representative value is calculated. , The contract forecast price is calculated based on the contract price actual representative value.

(19) In some embodiments, in the configuration of (16) above,
The approximate quantitative achievement acquisition unit acquires the approximate quantitative achievement information included in the record exceeding the preset threshold value of the similarity from the past achievement database.
When a plurality of the records have the same degree of similarity exceeding the threshold value, the representative value of the approximate quantitative result calculated by performing a predetermined calculation on the approximately quantitative actual set which is a set of the plurality of acquired approximately quantitative actual information is obtained. It is acquired as the said quantitative actual information.

According to the configurations of (17) to (19) above, when a plurality of records are acquired from the past record database as similar ones when calculating each similarity degree, each record information in the acquired plurality of records is used. On the other hand, the value obtained through calculation such as average may be used as the desired actual information. As a result, for example, even data located in the middle of a plurality of records stored in the past record database is corrected, so that it can be acquired as accurate record information.

According to at least one embodiment of the present invention, there is provided an electric power price prediction system that predicts a contract price in a variable market such as a one-day market or a same-day market.

It is a figure which shows roughly the structure of the electric power price prediction system which concerns on one Embodiment of this invention. It is a figure which shows the record (past record) managed by the past record database which concerns on one Embodiment of this invention. It is a figure which shows the calculation logic of the buy bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the calculation logic of the selling bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the other calculation logic of the buy bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the prediction logic which predicts the market breakdown of the buy bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the prediction logic which predicts the market breakdown of the selling bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the other prediction logic which predicts the market breakdown of the selling bid contract quantitative prediction which concerns on one Embodiment of this invention. It is a figure which shows the calculation logic of the area predicted price which concerns on one Embodiment of this invention. It is an example of the screen display which maps and shows the information of the inter-regional interconnection line including the possibility of market division which concerns on one Embodiment of this invention on a map. It is a figure which shows the calculation logic which calculates the contract forecast price in the one day before market which concerns on one Embodiment of this invention. It is a figure which shows the calculation logic which calculates the contract forecast price in the market on the day which concerns on one Embodiment of this invention. It is a figure which shows the balance adjustment of the bidding to the one day before market and the same day market which concerns on one Embodiment of this invention. It is a figure which shows the electric power price prediction system which cooperates with the power plant which concerns on one Embodiment of this invention.

Hereinafter, some embodiments of the present invention will be described with reference to the accompanying drawings. However, the contents described as embodiments or shown in the drawings are not intended to limit the scope of the present invention to this, and are merely explanatory examples. On the other hand, the expressions "equipped", "equipped", "equipped", "included", or "have" one component are exclusive expressions that exclude the existence of other components. No.

FIG. 1 is a diagram schematically showing a configuration of an electric power price prediction system 1 according to an embodiment of the present invention. In the electric power price prediction system 1, the market price of electric power delivered from the electric power selling business operator 91 to the electric power purchasing business operator 92 (retail business operator) at a desired (arbitrary) delivery date and time (target delivery date and time) fluctuates depending on the bidding situation. It is a system (device) that can calculate the contract forecast price Pc in the variable market (described above) for each bidding area (described later). The electric power selling business operator 91 is a business operator that sells electric power to another business operator, and includes a power generation business operator that owns a power plant 91p and a business operator that resells the purchased electric power to another business operator. The electric power purchasing business operator 92 is a business operator that sells electric power to a contracted consumer 93 who consumes electric power. The variable market means a market in which the contracted price of electric power fluctuates according to the bidding situation. For example, the Japan Electric Power Exchange (JEPX) is a one-day market (mentioned above) or a day market (mentioned above). To briefly explain the Japan Electric Power Exchange, as shown in FIG. 1, the power selling business 91 and the power purchasing business 92 access the power trading system 9 (computer system) via the Internet to trade power. A bid transaction for electric power is performed on the system 9.

More specifically, the Japan Electric Power Exchange divides the day into units of electricity measurement (0 to 30 minutes per hour, 30 minutes to 60 minutes), and 48 pieces (24H x 2) per day. Electric power products are available. Therefore, the electric power selling business operator 91 and the electric power purchasing business operator 92 respectively have an electric power product (time zone in units of 30 minutes), an electric power unit price (hereinafter, appropriately referred to as a price), and an electric power amount (hereinafter, appropriately, appropriately). By specifying (simply called quantity), bids (sell bid, buy bid) such as buying a specified amount if the price is less than the specified price or selling the specified amount in the specified price range are performed on the electric power trading system. In addition, when bidding, bid areas that are divided into areas such as Hokkaido are also specified in advance, and the bidding conditions are which area to sell the generated power or which area to buy the power to use. It is designed to be included in. Then, the transaction is executed when the bidding conditions (bid area, electric power product (time zone), electric power unit price, electric energy amount) of the selling bid by the electric power selling business operator 91 and the bidding conditions by the electric power purchasing business operator 92 match. do. The contract result (amount and price at which the sale was completed) is notified to the bidder after the contract processing, and then the clearing settlement (payment of purchase price, receipt of sale price) is performed. In the following, the power unit price, the contracted power unit price, and the contracted electric energy of the bidding conditions will be referred to as the bid price, the contracted price, and the contracted electric energy (contracted electric energy), respectively.

In this way, in the bidding transaction, the bid is executed when the bid conditions of the buy bid and the sell bid are met. Therefore, for the power seller 91, the higher the sell bid price, the higher the profit can be expected. If there is no buy bid that meets the bid conditions of the sell bid, the surplus power generation capacity will be left (equipment will be wasted), or the generated power cannot be sold. On the contrary, for the electric power purchaser 92, the lower the bid bid price, the higher the profit can be expected. Unable to buy power to supply. Therefore, the electric power price prediction system 1 calculates and presents the contract predicted price Pc so that the power selling business 91 and the power purchasing business 92 can determine an appropriate bid price.

Hereinafter, the electric power price prediction system 1 will be described in detail with reference to FIGS. 1 to 13.
FIG. 2 is a diagram showing a record (past record R) managed by the past record database Dr according to the embodiment of the present invention.

The electric power price prediction system 1 is composed of a computer. That is, it is equipped with a CPU (processor) (not shown), a memory such as ROM or RAM, or a storage device M such as an external storage device, and the CPU operates according to an instruction of a program (power price prediction program) loaded in the main storage device ( By performing data calculation, etc.), each functional part described later is realized. Further, the electric power price prediction system 1 is connected to an external output device such as a display or a printer, so that the contract predicted price Pc can be obtained via the external output device by the user of this system (power sales company 91 or power purchase business). Present to person 92).

Then, as shown in FIG. 1, the electric power price prediction system 1 includes a past record database Dr, a forward contract total amount acquisition unit 2, an environmental forecast acquisition unit 3, a quantitative result acquisition unit 4, and a quantitative forecast calculation. A unit 5 and a contract forecast price calculation unit 6 are provided, and a contract forecast price Pc of electric power traded in a variable market is calculated for a desired delivery date and time. Each of the above-mentioned functional parts included in the electric power price prediction system 1 will be described.

In the following description, the contracted estimated price in the desired bid area (hereinafter referred to as the target bid area) of the desired delivery date and time (hereinafter referred to as the target delivery date and time) in the variable market of the Japan Electric Power Exchange (JEPX). The case of calculating Pc will be described as an example. The target delivery date and time and the target bidding area are designated in advance by the business operator (91, 92) using the electric power price prediction system 1 of the present invention. At this time, the target bidding area may include one or more bidding areas, and the target delivery date and time may include one or more times on the desired delivery date (time zone corresponding to 48 electric power products; the same shall apply hereinafter). When such a plurality of bidding areas and a plurality of times are included, the contract predicted price Pc is calculated for each bidding area and each time (by electric power product). In addition, the variable market will be at least one of the one-day market and the same-day market.

Further, in the following description, the acquisition of information (data) from the past record database Dr performed by various functional units described later is performed by transmitting the search condition to the past record database Dr by, for example, SQL, and the information matching the search condition. Will be described as an example in the case of performing by receiving. At this time, in the comparison between the past record database Dr and the search conditions, it is determined whether or not the two match if the difference between the two is within a predetermined value (± α% or less, etc.) for each information field. For example, it does not have to be exactly the same, and it may be determined with a predetermined width to acquire the information for which the match is determined.

However, the present invention is not limited to the present embodiment. In some other embodiments, the acquisition of information (data) from the past record database Dr targets at least some of the records of the plurality of past record R (described later) stored in the past record database Dr. , The acquisition condition including the environmental forecast information Ec (described later) of the target delivery date and time in the target bidding area, and the information including the environmental actual information Er (described later) stored in the past record database Dr (information corresponding to the acquisition condition). The similarity is obtained for each record, and the approximate quantitative performance information Qr (described later) corresponding to the time of the target delivery date and time in the target bid area is acquired from the past performance database Dr based on the similarity, for example, those with high similarity. You may. More specifically, a known method that can obtain the similarity such as matching by dynamic programming is applied, or machine learning using various information stored in the past record R as a parameter is executed to compare. The information having the highest degree of similarity may be determined as a match through the determination of the match with (acquisition condition), and the information for which the match has been determined may be acquired. In addition, the threshold value of similarity is set in advance, and when a plurality of records exceeding the value are obtained, the value obtained by collecting the approximate quantitative performance information QR corresponding to each record and calculating the average or the like is obtained. It may be treated as performance information QR. The application destination of machine learning is not limited to the extraction of similar data from the above-mentioned past performance database Dr, and may be used for various predictions such as contracted amount, contract price, power generation capacity, market breakdown, and market division, which will be described later. .. In machine learning, various achievement information stored in the past achievement database Dr may be used as teacher data to determine a match in advance and to train various predictions associated with the determination to complete the learning model. The learning model may be trained (updated) while repeatedly comparing with the match judgment, prediction, and actual result (result). As a result, information with a high degree of similarity can be obtained from the past record database Dr.

The past performance database Dr is a database for storing the past performance information required for calculating the contract forecast price Pc. More specifically, as shown in FIG. 2, the past performance database Dr associates the environmental performance information Er, the approximately quantitative performance information QR, and the contract price performance information Pr with each other, and sorts them by delivery date and time. And memorize by bid area. The above-mentioned environmental performance information Er is a past performance of environmental information including weather information that affects electric power demand. More specifically, for example, in addition to weather, temperature, humidity, etc., information such as wind speed, amount of solar radiation, sunshine duration, time, and season may be included. Approximate quantitative performance information Qr is the past traded in each of the variable market where the electric power delivered on the past delivery date and time was traded and the forward market (described above) where the electric power was traded before the trading date of this variable market. It is the actual information of the total amount of electric power of. The contract price actual information Pr is the actual contract price of electric power traded in the past in the variable market.

Therefore, as shown in FIG. 2, the records managed by the past record database Dr (past record R) include date and time information, area information, environmental record information Er, about quantitative record information QR, and contract price record information. It has an information field for storing past performance information including Pr. That is, the date and time information that can specify the day (year, month, day) and the time indicates which past record R is the actual information of which time on which delivery date, and the area information indicates that each past record R is. Indicates in which bidding area the performance information is. In the area information, an area equivalent to the bidding area may be stored, or a constituent area that at least partially constitutes the bidding area may be stored. This constituent area may be an area of a unit of environmental forecast. Then, each past record R is generated by acquiring information to be stored in each information field from various databases described later and linking them with date / time information and area information.

The forward contract total amount acquisition unit 2 acquires the forward contract total amount Qm, which is the amount of electric power delivered in the target bid area on the target delivery date and time traded in the forward market. As already explained, the electric power on each delivery date is traded first in the forward market until the day before the day before the delivery date, but the forward contract total amount acquisition unit 2 tries to calculate the contract forecast price Pc. Obtain the total amount of forward contracts Qm for each delivery date and time by acquiring the contract result of the forward market for the target delivery date (target delivery date) at an appropriate timing such as after the deadline of the forward market for the target delivery date. do. In some embodiments, the electric power price prediction system 1 includes a forward information database Dm that automatically captures and stores the execution results of the forward market published by a publisher such as JEPX, and the forward execution is executed. The total amount acquisition unit 2 may acquire the forward contract total amount Qm from the forward information database Dm.

The environmental forecast acquisition unit 3 acquires the environmental forecast information Ec of the target delivery date and time in the target bidding area. The environmental forecast information Ec is the forecast of the above-mentioned environmental information, and the environmental forecast acquisition unit 3 acquires the environmental forecast information Ec of the target delivery date and time of the target bidding area. More specifically, the environmental forecast acquisition unit 3 may acquire the target environmental forecast information Ec from the environmental information database De that manages the environmental forecast information Ec for each predetermined area and time (see the figure to be described later). 3 to 4).

The approximately quantitative result acquisition unit 4 is based on a comparison between the environmental forecast information Ec of the target delivery date and time in the target bid area acquired by the environmental forecast acquisition unit 3 and the environmental record information Er stored in the past record database Dr. Then, the approximate quantitative performance information Qr corresponding to the time of the target delivery date and time in the target bidding area is acquired from the past performance database Dr. That is, the approximately quantitative result acquisition unit 4 searches the past result database Dr under the search conditions including the environmental forecast information Ec of the target delivery date and time in the target bid area. Specifically, for example, in the embodiment shown in FIG. 1, the environmental forecast information Ec is compared with the environmental actual information Er of each past actual record R managed by the past actual database Dr, and is consistent with the environmental forecast information Ec (same). Or similar. The same applies hereinafter.) From one or more past record R having the environmental record information Er, the past record R in which the target bidding area and the target delivery date and time match is extracted. In some other embodiments, the past performance database Dr is searched using the target bid area, the target delivery date and time, and the environmental forecast information Ec as search conditions, and the area information, the date and time information time, and the time of the date and time information that match the search conditions are searched. The past record R having the environmental record information Er may be extracted. The date and time information used for extracting the past record R may be day and time. When the target delivery date and time includes a plurality of time zones such as one day, the past record R close to the tendency of the transition of the environmental forecast information in the plurality of time zones may be extracted. Then, the approximately quantitative actual information Qr possessed by the extracted past actual record R is acquired.

In addition, when a plurality of past record records R are extracted from the past record database Dr, the approximately quantitative result acquisition unit 4 selects one of them and performs about. Quantitative actual information Qr may be acquired, or a value calculated by performing a predetermined calculation such as an average on a plurality of about quantitative actual information Qr (about quantitative actual set) obtained from them (representative of about quantitative actual result). Value) may be acquired as about quantitative performance information Qr. In addition, period environmental information (for example, one week including the delivery date and time) obtained by adding the environmental record information Er for a certain period up to the target delivery date and time to the environmental forecast information Ec of the target delivery date and time in the target bid area, and the past record database. Based on the comparison between the above-mentioned period environment information stored in Dr and the environmental record information Er for the same period (for example, one week), the past record database Dr corresponds to the time of the target delivery date and time in the target bid area. Approximate quantitative performance information Qr may be acquired. As a result, it is possible to extract a past record R having a similar tendency to capture the transition of environmental changes for a certain period including the target delivery date and time.

The approx. Quantitative forecast calculation unit 5 subtracts the forward contract total amount Qm from the approx. Quantitative actual information Qr acquired by the approx. Calculate Qc. In a variable market, the power delivered on each delivery date and time (48 each time zone of each delivery date) is the sum of the power traded in both the forward and variable markets with respect to that delivery date and time. The amount of electric power other than the electric power contracted in the forward market in the contracted quantitative performance information Qr is traded.

Therefore, in the present invention, in the approximately quantitative prediction calculation unit 5, the approximately quantitative result acquisition unit 4 acquires the target delivery date and time from the approximately quantitative result information Qr corresponding to the target delivery date and time. The electric energy (= Qr−Qm) calculated by subtracting the total amount Qm is defined as the approximately quantitative prediction Qc. That is, the electric power price prediction system 1 of the present invention equals the amount of electric power expected to be executed with respect to the target delivery date and time in the target bidding area to that of the past actual results (approx. Quantitative actual information Qr) that match from the viewpoint of the environmental forecast information Ec. Predict as. Based on this forecast, by subtracting the electric energy contracted earlier in the forward market (total amount of forward contracts Qm) with respect to the target delivery date and time in the target bid area from this predicted amount (approx. Quantitative actual information Qr). , Predict the expected electric energy (approx. Quantitative forecast Qc) to be executed in the variable market with respect to the target delivery date and time in the target bidding area.

The contract price forecast calculation unit 6 acquires the contract price actual information Pr from the past actual database Dr based on the contract quantitative forecast Qc calculated by the contract quantitative forecast calculation unit 5, and based on the acquired contract price actual information Pr. The contract forecast price Pc is calculated. More specifically, the past performance database Dr associates the contract price performance information Pr with the contracted price forecast calculation unit 5 and stores the contracted price forecast Qcr, which is the contracted price forecast Qc calculated in the past. It has become. Then, the contract price forecast price calculation unit 6 acquires the contract price actual information Pr from the past actual database Dr based on the comparison between the contract quantitative forecast Qc and the contract quantitative forecast actual Qcr stored in the past actual database Dr.

That is, as shown in FIG. 2, the past actual record R has an information field for storing the approximately quantitative forecast actual Qcr, so that the past actual database Dr has the past approximately quantitative forecast in the variable market. The relationship between (contract quantitative forecast actual Qcr) and the actual contract price (contract price actual information Pr) with respect to the forecast can be understood. Therefore, the contracted price actual information Pr of the past actual record R having the approximately quantitative predicted actual Qcr that matches the approximately quantitative predicted Qc is acquired as being similar to the contracted predicted price Pc to be obtained. In some other embodiments, the contract forecast price calculation unit 6 uses the contract price from the past record database Dr based on the comparison between the contract forecast Qc and the contract quantitative actual information Qr stored in the past record database Dr. Achievement information Pr may be acquired. As a result, the contract price actual information Pr used for calculating the contract estimated price Pc can be appropriately acquired from the past actual database Dr, and the contract estimated price Pc can be calculated. In the embodiment shown in FIG. 1, as shown in FIG. 1, necessary information including the contract predicted price Pc is transmitted to the power selling business operator 91 and the power purchasing business operator via the output unit 14 included in the electric power price prediction system 1. It is presented to an output device (not shown) such as a display included in the 92 or the like.

When the contract forecast price calculation unit 6 acquires a plurality of contract price actual information Pr from the past actual database Dr, the contract forecast price calculation unit 6 may calculate the contract forecast price based on any one of them. , Calculate the contract forecast price based on the value (contract price actual representative value) calculated by performing a predetermined calculation such as average for multiple contract quantitative actual information Qr (contract price actual set) obtained from them. You may. Further, in some other embodiments, when a plurality of past actual record R is acquired by the approximately quantitative result acquisition unit 4, the approximately quantitative prediction calculation unit 5 makes a forward contract from the plurality of approximately quantitative actual information Qr. A plurality of about quantitative prediction Qc are acquired by subtracting the total amount Qm, and the contract prediction price calculation unit 6 is a past performance database based on each of the plurality of about quantitative prediction Qc calculated by the about quantitative prediction calculation unit 5. Multiple contract price actual information such as acquiring multiple contract price actual information Pr from Dr and calculating the average value etc. of the acquired multiple contract price actual information Pr (contract price actual information) (contract price actual representative value). The contract forecast price Pc may be calculated based on Pr.

According to the above configuration, the electric power price prediction system 1 calculates the contract forecast price Pc of the variable market at the desired delivery date and time (target delivery date and time) to be predicted based on the past record information stored in the past record database Dr. Predict. Specifically, the approximate quantitative actual information Qr associated with the environmental actual information Er that matches (same or similar) the environmental forecast information Ec of the target delivery date and time is used as the predicted electric energy to be delivered on the target delivery date and time. In addition, by subtracting the amount of power traded in the forward market (total amount of forward contracts Qm) from the predicted power amount with respect to the target delivery date and time, the amount of power traded (contracted) in the variable market is predicted. ..

Then, for example, the result of the approximately quantitative prediction Qc calculated in the past (approx. Quantitative prediction actual Qcr) or the approximately quantitative actual information Qr is associated with the past contract price actual information Pr stored in the past actual record database Dr. If it is memorized, the contracted price actual information Pr corresponding to the approximately quantitative forecast Qc or the contracted price actual information Qr corresponding to the contracted quantitative forecast Qc is acquired and used as the contracted predicted price Pc of the target delivery date and time. For example, the contract estimated price Pc of the target delivery date and time is calculated based on the contract price actual information Pr. In this way, the contract forecast price Pc at the desired delivery date and time can be accurately predicted based on the past results. In addition, by making it possible to calculate the contract forecast price Pc with high accuracy, it is possible to present the optimum buying and selling bid in the variable market.

In some embodiments, as shown in FIG. 1, the calculation of the approximately quantitative prediction Qc may be performed separately for each buy bid and sell bid. More specifically, in some embodiments, as shown in FIG. 1, the electric power price prediction system 1 includes a fuel price acquisition unit 44 that acquires fuel price information Fc of the fuel used for power generation by the electric power seller 91, and a fuel price acquisition unit 44. Further, the power generation capacity prediction acquisition unit 45 for acquiring the power generation capacity prediction Ac at the target delivery date and time of the power generation plant 91p owned by the power sales company 91 is provided. Further, as shown in FIG. 2, the past performance database Dr includes environmental performance information Er, about quantitative performance information Qr, contract price performance information Pr, and power generation capacity performance information for each past date and time in the power seller 91. Ar and the fuel price actual information Fr, which is the past fuel price information Fc, are stored in association with each other. In the embodiment shown in FIG. 2, the business ID that identifies the power sales business 91 indicates which power sales business 91 the power generation capacity actual information Ar belongs to.

Further, as shown in FIG. 1, the above-mentioned about quantitative result acquisition unit 4 has a buy bid contract quantitative result acquisition unit 4b and a sell bid contract quantitative result acquisition unit 4s. Based on the comparison between the environmental forecast information Ec and the environmental performance information Er, the purchase bid contracted quantitative result acquisition unit 4b acquires the contracted quantitative performance information Qr from the past record database Dr and obtains the bid bid contracted quantitative result Qrb. On the other hand, in addition to the comparison between the environmental forecast information Ec and the environmental performance information Er, the selling bid contract quantitative result acquisition unit 4s compares the power generation capacity prediction Ac with the power generation capacity performance information Ar, and the fuel price information Fc and the fuel price. Based on the comparison with the actual information Fr, the approximate quantitative actual information Qr is acquired from the past actual database Dr and used as the selling bid about quantitative actual Qrs.

Further, the above-mentioned about quantitative prediction calculation unit 5 has a buy bid about quantitative prediction calculation unit 5b, a sell bid about quantitative prediction calculation unit, and 5s. The bid bid contract quantitative forecast calculation unit 5b calculates the bid bid contract quantitative forecast Qcb by subtracting the forward contract total amount Qm from the bid bid contract quantitative forecast calculation unit Qrb. On the other hand, the sell bid contract quantitative prediction calculation unit 5s calculates the sell bid contract quantitative forecast Qcs by subtracting the forward contract total amount Qm from the sell bid contract quantitative forecast calculation unit Qrs. More specifically, the buy bid contract quantitative forecast calculation unit 5b is connected to the bid bid contract quantitative result acquisition unit 4b and the forward contract total amount acquisition unit 2, and the bid bid acquired by the bid bid contract quantitative result acquisition unit 4b. The contracted total amount Qrb and the forward contracted total amount Qm acquired by the forward contracted total amount acquisition unit 2 are received, and the bid bid contracted total amount predicted Qcb is calculated as described above (Qcb = Qrb-Qm). Similarly, the sell bid contract quantitative forecast calculation unit 5s is connected to the sell bid contract quantitative result acquisition unit 4s and the forward contract total amount acquisition unit 2, and the sell bid contract quantitative result acquisition unit 4s has acquired the sell bid contract quantitative. The actual Qrs and the forward contract total amount Qm acquired by the forward contract total amount acquisition unit 2 are received, and the sell bid contract quantitative prediction Qcs is calculated as described above (Qcs = Qrs-Qm).

The calculation logic of the buy bid contract quantitative prediction Qcb described above will be described with reference to FIG. FIG. 3 is a diagram showing a calculation logic of a bid bid contract quantitative prediction Qcb according to an embodiment of the present invention. As shown in FIG. 3, in step S31, the environmental forecast information Ec is acquired by the environmental forecast acquisition unit 3 (S31a), and the forward contract total amount Qm is acquired by the forward contract total amount acquisition unit 2 (S31b). In step S32, the purchase bid contract quantitative result acquisition unit 4b compares the environmental forecast information Ec acquired in step S31 with the environmental record information Er of each past record R managed by the past record database Dr, and makes an environmental forecast. One or more past record R having the environmental record information Er that matches the information Ec is extracted. Then, in step S33, as described above, the buy bid contract quantitative result acquisition unit 4b sets, for example, the bid bid contract quantitative result Qr of the extracted past record record R as the buy bid contract quantitative result Qr. After that, in step 34, the bid bid contract quantitative prediction calculation unit 5b calculates the bid bid contract quantitative forecast Qcb by subtracting the forward contract total amount Qm from the bid bid contract quantitative forecast Qrb (Qcb = Qrb-Qm). ..

Similarly, the calculation logic of the above-mentioned sell bid contract quantitative prediction Qcs will be described with reference to FIG. FIG. 4 is a diagram showing a calculation logic of a bid contract quantitative prediction Qcs according to an embodiment of the present invention. Steps S40 (S40a to S40c) of FIG. 4 will be described later. As shown in FIG. 4, after step S40, in step S41, the environmental forecast information Ec is acquired by the environmental forecast acquisition unit 3 (S41a), and the forward contract total amount Qm is acquired by the forward contract total amount acquisition unit 2 (S41a). S41b), the fuel price information Fc is acquired by the fuel price acquisition unit 44 (S41c), and the power generation capacity forecast Ac is acquired by the power generation capacity forecast acquisition unit 45 (S41d). In step S42, the selling bid contract quantitative result acquisition unit 4s determines the environment of the environmental forecast information Ec, the power generation capacity prediction Ac, and the fuel price information Fc acquired in step S41, and the environment of each past record R managed by the past record database Dr. Comparing the performance information Er, the power generation capacity performance information Ar, and the fuel price performance information Fr, respectively, the environmental performance information Er, the power generation capacity performance information Ar, and the fuel that match the environmental forecast information Ec, the power generation capacity prediction Ac, and the fuel price information Fc. Extract one or more past performance records R having price performance information Fr. Then, in step S43, as described above, the selling / bidding contracted quantitative result acquisition unit 4s sets, for example, the approximate quantitative result information QR of the extracted past record R as the selling / bidding contracted quantitative result Qrs. After that, in step 34, the sell bid contract quantitative prediction calculation unit 5s calculates the sell bid contract quantitative prediction Qcs by subtracting the forward contract total amount Qm from the sell bid contract quantitative forecast Qrs (Qcs = Qrs-Qm). ..

Then, when the calculation of the approximate quantitative prediction Qc is performed separately for each of the buy bid and the sell bid in this way, each of the calculated buy bid contract quantitative forecast Qcb and the sell bid contract quantitative forecast Qcs calculates the contract forecast price. It is input to the part 6. Therefore, when the calculation of the approximate quantitative forecast Qc is performed separately for each bid and sell bid in this way, the contract forecast price calculation unit 6 includes the buy bid contract quantitative forecast Qcb and the sell bid contract quantitative forecast Qcs. Obtain the contract price actual information Pr from the past actual database Dr based on the comparison between the buy bid contract quantitative forecast actual Qcr (b) and the sell bid contract quantitative forecast actual Qcr (s) stored in the past actual database Dr. ..

According to the above configuration, the bid bid contract quantitative forecast Qcb is the contract quantitative actual information Qr when the environmental performance information Er (past) matches the environmental forecast information Ec of the target delivery date and time. In other words, it corresponds to the amount of power demand on the target delivery date and time predicted based on environmental factors. Therefore, the bid bid contract quantitative forecast Qcb corresponds to the electricity demand forecast in the variable market. On the other hand, in the sell bid contract quantitative forecast Qcs, each of the environmental performance information Er (past), the power generation capacity performance information Ar (past), and the fuel price performance information Fr (past) is the environmental forecast information Ec of the target delivery date and time, respectively. It is the approximate quantitative actual information Qr when the power generation capacity forecast Ac and the fuel price information Fc match or comprehensively match. In other words, it corresponds to the amount of power supply on the delivery date predicted based on environmental factors and power generation circumstances factors. Therefore, the sell bid contract quantitative forecast Qcs corresponds to the power supply forecast in the variable market.

Therefore, the contract forecast price calculation unit 6 determines the contract forecast price Pc when the power demand forecast and the power supply forecast of the target delivery date and time in the target bid area are the buy bid contract quantitative forecast Qcc and the sell bid contract quantitative forecast Qcs. , The contract price actual information of the past actual record R having the buy bid contract quantitative forecast Qcb and the sell bid contract quantitative forecast Qcr (b) and the sell bid contract quantitative forecast Qcr (s). Calculated based on Pr. In this way, by referring to the past record database Dr, the power demand forecast (buy bid contract quantitative forecast result Qcr (b)) and the power supply amount forecast (sell bid contract quantitative forecast result Qcr (s)) performed in the past ))), By obtaining the contract price actual information Pr actually executed, the accuracy of the contract predicted price Pc calculated based on the contract price actual information Pr can be further improved.

In some embodiments, the power generation capacity prediction Ac described above may be calculated based on power source configuration information and power plant 91p shutdown information. Specifically, as shown in FIG. 4, the electric power price prediction system 1 further includes a power source information database Dp and a power generation capacity calculation unit 46.
The power supply information database Dp is a database that stores the power supply configuration information and the stop information of the power plant 91p, which are the information of the equipment constituting the power plant 91p owned by the power sales company 91, for each power sales company. The power source configuration information is information such as the type of power generation equipment (equipment) such as a BTG (Boiler Turbine Generator) and a gas engine, and specifications of power generation capacity. Further, the stop information of the power plant 91p is classified into a planned stop and a trouble stop. Information on planned outages and trouble outages will be provided by posting on the power generation information disclosure site for each power seller 91. The power source configuration information and the stop information of the power plant 91p are stored in the power supply information database Dp based on the notification from each power sales company 91.

The power generation capacity calculation unit 46 calculates the power generation capacity of the power sales company 91 based on the power generation configuration information and stop information of the power plant 91p described above. More specifically, for example, the maximum power generation capacity (kW, MW, etc.) that can be calculated based on the power generation configuration information of the power plant 91p is calculated and used as the power generation capacity. In addition, if there is a planned stop or trouble stop on the target delivery date and time, the power generation capacity is calculated by subtracting the amount that cannot be generated due to the stop of the power plant from the stop information of the power generation plant 91p in advance from the maximum power generation capacity. do.
Then, the power generation capacity prediction acquisition unit 45 described above acquires the power generation capacity calculated by the power generation capacity calculation unit 46 as the power generation capacity prediction Ac.

According to the above configuration, the power generation capacity prediction Ac of the target delivery date and time in the target bidding area can be appropriately calculated based on the power generation configuration information and the stop information of the power plant 91p. Therefore, it is possible to obtain the selling bid contract quantitative actual Qrs with higher accuracy.

Further, in some embodiments, as shown in FIG. 3, the bid bid contract quantitative prediction calculation unit 5b may correct the bid bid contract quantitative prediction Qcb.
Specifically, as shown in FIG. 2, the past performance database Dr is a bid bid contract quantitative forecast that is a bid bid contract quantitative forecast Qcb calculated based on the bid bid contract quantitative forecast Qrb as described above in the past. The actual Qcr (b) is stored in association with the approximately quantitative actual information Qr, which is the purchased bid contract quantitative forecast actual Qcr (b) used for calculating the bid bid contract quantitative forecast actual Qcr (b).

Further, the electric power price prediction system 1 includes the buy bid contract quantitative forecast Qcb calculated by the buy bid contract quantitative forecast calculation unit 5b described above and the buy bid contract quantitative forecast result Qcr (b) stored in the past record database Dr. Based on the comparison between Further, a buy-bid contract quantitative prediction error calculation unit 47b for calculating a buy-bid contract quantitative prediction error based on Qr is further provided. In the embodiment shown in FIG. 3, among a plurality of past performance records R acquired by searching the past performance database Dr by the environmental forecast information Ec, the buy bid contract quantitative prediction Qcb, which is a search condition, and the past performance database Dr are managed. Compared with the buy bid contract quantitative forecast result Qcr (b) of each past record R, the past record extracted as having the buy bid contract quantitative forecast result Qcr (b) that matches the buy bid contract quantitative forecast Qcb. From the record R, the bid contract quantitative forecast result Qcr (b) and the contract quantitative record information Qr are acquired as a set. Then, for example, the error rate is calculated as {Qr-Qcr (b)} / Qr as the bid bid contract quantitative prediction error, and this is multiplied by the previously calculated bid bid contract quantitative prediction Qcb to correct the error. The later bid bid contract quantitative prediction Qcb is calculated.

Then, the above-mentioned bid bid contract quantitative prediction calculation unit 5b calculates the bid bid contract quantitative prediction Qcb using the bid bid contract quantitative prediction error calculated by the bid bid contract quantitative prediction error calculation unit 47b. For example, the bid bid contract quantitative prediction calculation unit 5b calculates the corrected bid bid contract quantitative forecast Qcb by multiplying the bid bid contract quantitative forecast Qcb calculated as described above by the bid bid contract quantitative forecast error. do.

According to the above configuration, the buy bid contract quantitative prediction Qcb is corrected by the buy bid contract quantitative prediction error. Thereby, the accuracy of the bid bid contract quantitative prediction Qcb can be improved.

Similarly, in some embodiments, as shown in FIG. 4, the sell / bid contract quantitative prediction calculation unit 5s may correct the sell / bid contract quantitative prediction Qcs.
Specifically, as shown in FIG. 2, the past record database Dr is a sell bid contract quantitative forecast Qcs, which is a sell bid contract quantitative forecast Qcs calculated based on the sell bid contract quantitative result Qrs as described above in the past. The actual Qcr (s) is stored in association with the approximate quantitative actual information Qr, which is the sold bid contract quantitative forecast actual Qcr (s) used for calculating the sold bid contract quantitative forecast actual Qcr (s).

Further, the electric power price prediction system 1 includes the sell bid contract quantitative forecast Qcs calculated by the sell bid contract quantitative forecast calculation unit 5s described above and the sell bid contract quantitative forecast result Qcr (s) stored in the past record database Dr. Based on the comparison between Further, a sell-bid contract quantitative prediction error calculation unit 47s for calculating a sell-bid contract quantitative prediction error based on Qr is further provided. In the embodiment shown in FIG. 4, among a plurality of past performance records R acquired by searching the past performance database Dr using the environmental forecast information Ec, the search condition, the sell bid contract quantitative prediction Qcs, and the past performance database Dr are managed. The past performance extracted as having the selling bid contract quantitative forecasting result Qcr (s) that matches the selling bid contract quantitative forecast Qcs by comparing with the selling bid contract quantitative forecasting performance Qcr (s) of each past record R. From the record R, the bid contract quantitative forecast result Qcr (s) and the contract quantitative record information Qr are acquired as a set. Then, for example, the error rate is calculated as {Qr-Qcr (s)} / Qr as the sell bid contract quantitative prediction error, and this is multiplied by the previously calculated sell bid contract quantitative prediction Qcs to correct the error. Calculate the later sell bid contract quantitative forecast Qcs.

Then, the above-mentioned sell bid contract quantitative prediction calculation unit 5s calculates the sell bid contract quantitative prediction Qcs by using the sell bid contract quantitative prediction error calculated by the sell bid contract quantitative prediction error calculation unit 47s. For example, the sell / bid contract quantitative prediction calculation unit 5s calculates the corrected sell bid contract quantitative forecast Qcs by multiplying the sell bid contract quantitative forecast Qcs calculated as described above by the sell bid contract quantitative forecast error. do.

According to the above configuration, the sell bid contract quantitative prediction Qcs is corrected by the sell bid contract quantitative prediction error. As a result, the accuracy of the sell / bid contract quantitative prediction Qcs can be improved.

FIG. 5 is a diagram showing another calculation logic of the bid bid contract quantitative prediction Qcb according to the embodiment of the present invention. In the above-described embodiment, the bid bid contract quantitative prediction Qcb for the entire bidding area is calculated at once. In some other embodiments, as shown in FIG. 5, the power demand for each contracted customer 93 of the power supply of the power purchaser 92 is calculated based on the past performance database Dr, and the purchase in the target bidding area is performed. The bid contract quantitative prediction Qcb may be calculated. The electric power demand is the sum of the demands of the contracted customer 93 (transmission destination), but since the demand tendency according to the environmental forecast information Ec may differ for each contracted customer 93, the contractor Demand is forecast for each consumer 93.

Specifically, the electric power price prediction system 1 provides transmission destination information C (contractor consumer information) including the location of the transmission destination regarding the transmission destination (contractor consumer 93) to which the power purchaser 92 transmits power. Further, a power transmission destination information database Dc for storing and a power transmission destination information acquisition unit 48 for acquiring power transmission destination information C of a plurality of power transmission destinations located in the target bidding area from the power transmission destination information database Dc are further provided. Then, the environmental forecast acquisition unit 3 acquires the environmental forecast information Ec for each of the locations of the plurality of power transmission destinations acquired by the power transmission destination information acquisition unit 48. That is, the environmental forecast information Ec of the constituent area where the location of the power transmission destination exists is acquired. The environmental forecast information Ec includes the results of weather forecasts for each city, ward, town, and village, and for each predetermined area such as the eastern and western parts of a certain area. In this case, the constituent areas are predicted by the environmental forecast information Ec. It may match the unit of the area. Further, as described above, the buy / bid contract quantitative prediction calculation unit 5b is based on the comparison between each of the environmental forecast information Ec for each location of the transmission destination and the environmental actual information Er stored in the past actual database Dr. , One or more of the approximate quantitative actual information Qr is acquired from the past actual database Dr, and the total of the acquired approximate quantitative actual information Qr for each location of the transmission destination is used as the buy bid contract quantitative forecast Qcb.

In some embodiments, the above-mentioned power transmission destination information C includes the location, personnel composition (age, gender, etc.), housing type (detached house, collective, etc.), facility type (commercial facility, industrial facility, scale, etc.). ) Etc. may be included. For example, in detached houses, temperature and humidity have a relatively strong influence on electricity demand, commercial facilities have low electricity demand at night, and businesses (industrial facilities) usually have low electricity demand on holidays. Therefore, by including information other than the above-mentioned location in the information of the power transmission destination information C, it becomes possible to predict the electric power demand more accurately. In this case, the past record database Dr stores the power transmission destination information C in association with the business ID in the past record R, or the date and time of the above-mentioned date and time information and the business ID. The power transmission destination record information Cr, which is the result of the power transmission destination information C at the time of information, is associated and stored. Then, the buy bid contract quantitative prediction calculation unit 5b described above has the environmental forecast information Ec and the power transmission destination information C for each environmental forecast information Ec for each location of the power transmission destination, and the environmental performance information stored in the past performance database Dr. Based on the comparison with Er and the power transmission destination information C (power transmission destination performance information Cr), the approximate quantitative performance information Qr is acquired from the past performance database Dr, and the approximately quantitative performance information Qr for each location of the acquired power transmission destination is obtained. Let the total be the buy bid contract quantitative forecast Qcb. By including the power transmission destination information C in the past performance database Dr, it becomes difficult to obtain a completely matching past performance record R, but the past performance record R having information having a high degree of similarity to the power transmission destination information C is matched. You may obtain it as a thing to do.

The calculation logic of the buy bid contract quantitative prediction Qcb described above will be described according to the flow of FIG. In step S51, the power transmission destination information acquisition unit 48 acquires, for example, the power transmission destination information C of the power transmission destination located in the target bidding area. In the example of FIG. 2, since the past record R has the area information and the business operator ID, the area information is linked to the business operator ID by looking at the business operator ID of the record which is the target bidding area. Since the power transmission destination located in the target bidding area is known, appropriate power transmission destination information C is acquired. In step S52, the environmental forecast acquisition unit 3 acquires the environmental forecast information Ec for each of the locations of the plurality of power transmission destinations. In step S53, for each acquired environmental forecast information Ec, a search condition including the environmental forecast information Ec and an information field corresponding to the search condition stored in the past record database Dr (environmental record information Er or environmental record information). Compared with Er and the transmission destination information C), the corresponding past record R is extracted from the past record database Dr. Then, the approximate quantitative performance information Qr possessed by the extracted past performance record R is totaled, and totaled for each target bidding area and for each time zone corresponding to the electric power product.

According to the above configuration, for each power purchase business operator 92, the approximate quantitative result information QR is acquired for each power transmission destination (contractor) contracted for power supply, and the total is used as the purchase bid contract quantitative result Qrb. In this way, the calculation accuracy can be improved by accumulating the electric power demand for each contractor customer predicted based on the power transmission destination information C and calculating the final bid bid contract quantitative prediction Qcb.

Next, an embodiment of calculating the contract forecast price Pc for each market such as the one-day market and the same-day market included in the variable market will be described with reference to FIGS. 6 to 7. FIG. 6 is a diagram showing a prediction logic for predicting the market breakdown of the bid contract quantitative prediction Qcb according to the embodiment of the present invention. Further, FIG. 7 is a diagram showing a prediction logic for predicting the market breakdown of the bid contract quantitative prediction Qcs according to the embodiment of the present invention. As mentioned above, a variable market may include multiple markets, such as a same-day market that determines the market price on the delivery date and a one-day market that determines the market price one day before the delivery date. Yes, we predict how the approximately quantitative forecast Qc will be dispersed among the multiple markets that make up the variable market, by buy bid and sell bid.

Specifically, in some embodiments, the past performance database Dr is the environmental performance information Er, the buy bid contract quantitative forecast result Qcr (b), the sell bid contract quantitative forecast result Qcr (s), and the market on the day. Mutual correspondence between the current day's quantitative actual information Qrt, which is the amount of electricity traded in the past date and time, and the one-day's quantitative actual information Qrp, which is the amount of electricity traded in the market one day ago I remember it (see Fig. 2). In the embodiment shown in FIG. 2, the approximate quantitative results are stored separately for each market by associating the current day's approximately quantitative results information QRt, the one-day-before contracted quantitative results information QRp, and the forward contracted quantitative results information QRm. .. Further, as shown in FIG. 1, the electric power price prediction system 1 includes the bid / bid breakdown result acquisition unit 71 and the bid market breakdown forecast calculation unit 72 shown in FIG. 6, and the sell / bid breakdown record acquisition unit 73 shown in FIG. A market breakdown forecast calculation unit 7 (see FIG. 1) including a bid market breakdown forecast calculation unit 74 is further provided.
The above-mentioned functional units included in the market breakdown forecast calculation unit 7 will be described respectively.

As shown in FIG. 6, the bid / bid breakdown performance acquisition unit 71 includes the bid / bid contract quantitative forecast Qcb and the environmental forecast information Ec calculated by the bid / bid contract quantitative forecast calculation unit 5b, and the purchase stored in the past record database Dr. Based on the comparison between the bid contract quantitative forecast result Qcr (b) and the environmental record information Er, the bid contract quantitative forecast result Qrt and the one-day-ahead quantitative record information Qrp are acquired. In the embodiment shown in FIG. 6, one or more past record R is extracted by searching the past record database Dr using the buy bid contract quantitative prediction Qcb and the environmental forecast information Ec as search conditions, and the extracted past record R has. Acquire the daily quantitative actual information Qrt and the one-day previous quantitative actual information QRp. In addition, the bid market breakdown forecast calculation unit 72 is based on the same-day quantitative performance information QR and the one-day-a-day prior quantitative performance information QRp acquired by the bid breakdown performance acquisition unit 71, and the same-day market and one that constitute a variable market. Calculate the actual bid breakdown Drb, which is the breakdown of the amount of electricity traded in each of the day-to-day markets. Further, the buying bid market breakdown forecast calculation unit 72 sets the buying bid breakdown actual Drb as the buying bid market breakdown forecast Dcb, and based on the buying bid breakdown actual Drb, the same-day market and the one-day-ahead market in the buying bid contract quantitative forecast The breakdown of the bid market breakdown forecast Dcb is calculated. The bid breakdown record Drb may be calculated in advance and stored in the past record database Dr. In this case, the bid market breakdown forecast calculation unit 72 may use the past record database Dr to store the bid breakdown record Drb. It may be regarded as calculated by acquiring.

In the embodiment shown in FIG. 6, the bid / bid breakdown forecast calculation unit 72 is connected to the bid / bid breakdown result acquisition unit 71 to acquire the current day's approximate quantitative result information QR and the one day's previous day's approximate quantitative result information QRp. In addition, the bid bid market breakdown forecast calculation unit 72 calculates the ratio (buy bid breakdown result Drb) between the current day's approximate quantitative result information QRt and the one day's previous day's approximate quantitative result information QRp, and uses the calculated ratio as the buy bid breakdown result Drb. There is.

On the other hand, as shown in FIG. 7, the selling bid breakdown result acquisition unit 73 includes the selling bid contract quantitative forecast Qcs and the environmental forecast information Ec, and the selling bid contract quantitative forecast result Qcr (s) stored in the past record database Dr. And, based on the comparison with the environmental performance information Er, the current day's approximately quantitative performance information Qrt and the one day's previous day's approximately quantitative performance information Qrp are acquired. In the embodiment shown in FIG. 7, one or more past record R is extracted by searching the past record database Dr using the sell bid contract quantitative forecast Qcs and the environmental forecast information Ec as search conditions, and the extracted past record R has. Acquire the daily quantitative actual information Qrt and the one-day previous quantitative actual information QRp. In addition, the bid market breakdown forecast calculation unit 74 is based on the acquired current day's quantitative performance information QR and one day's previous day's quantitative performance information QRp, and the electric energy traded in each of the current day market and the one day's market that constitute the variable market. Calculate the actual sales / bid breakdown Drs, which is the breakdown of the quantity. Further, the selling bid market breakdown forecast calculation unit 74 sets the selling bid breakdown actual Drs to the selling bid market breakdown forecast Dcs, and based on the selling bid breakdown actual Drs, the same-day market and the one-day-ahead market in the selling bid contract quantitative forecast Qcb. The bid market breakdown forecast Dcs, which is the breakdown of the above, is calculated. The selling bid breakdown actual Drs may be calculated in advance and stored in the past actual database Dr. In this case, the selling bid market breakdown forecast calculation unit 74 may perform the selling bid breakdown actual Drs from the past actual database Dr. It may be regarded as calculated by acquiring.

Since the embodiment shown in FIG. 7 is the same as the embodiment described in FIG. 6 by replacing the buy bid with the sell bid, the description thereof will be omitted.
Then, the contract forecast price calculation unit 6 is based on at least one of the current day market and the one-day-ahead market based on the buy bid contract quantitative forecast Qbcb, the bid bid market breakdown forecast Dcb, the sell bid contract quantitative forecast Qcs, and the sell bid market breakdown forecast Dcs. (See FIGS. 10 to 11 described later).

In addition, when a plurality of past performance records R are extracted by the bid bid breakdown forecast calculation unit 72 or the sell bid market breakdown forecast calculation unit 74 by the bid bid breakdown performance acquisition unit 71 or the bid bid breakdown performance acquisition unit 73, , Any one of them may be selected, and a plurality of current day's approximately quantitative actual information Qrt (current day's approximately quantitative actual result set) and a plurality of one-day's previous approximately quantitative actual information Qrp (1) obtained from them. Value calculated by performing a predetermined calculation such as average for the previous day's approximate quantitative result set) (the current day's approximate quantitative representative value based on the current day's approximate quantitative result set, the one day's approximate quantitative result representative value based on the previous day's approximate quantitative result set) You may calculate the buy bid breakdown record Drb or the sell bid breakdown record Drs based on.

However, the present invention is not limited to the present embodiment. In some other embodiments, as shown in FIG. 8A, the above-mentioned bid market breakdown forecast Dcs may be calculated using the operable power supply information obtained from the power supply information database Dp. Specifically, the operable power source existing in the target bidding area is extracted based on the power source configuration information and the stop information of the power plant 91p stored in the power source information database Dp described above. In addition, the responsiveness to output fluctuations is evaluated for each operable power source based on the power source configuration information of each of the extracted operable power sources, and each operable power source is classified into a plurality of levels according to the quality of this responsiveness (the responsiveness is evaluated). Level of FIG. 8 (b). Further, the total output (total rated generator output in FIG. 8B) is calculated based on the power source configuration information for each response level of the operable power source classified in this way. Then, for each responsiveness level, the total output and the environmental forecast information Ec are compared with the total output (information field) and the environmental forecast information Ec of each past actual record R managed by the past actual database Dr, and the total output and the environment are compared. The past record R having the environmental record information Er that matches the forecast information Ec is extracted, and the sell / bid market breakdown forecast Dcs is calculated in the same manner as described above. As a result, it is possible to extract past results having a higher degree of similarity to the status of the target delivery date and time in the target bid area, so that the prediction accuracy can be improved.

According to the above configuration, the contract forecast price Pc can be calculated for each market such as the same-day market and the one-day-ahead market that constitute the variable market.

Further, in some embodiments, as shown in FIGS. 1 and 9A, the electric energy price forecasting system 1 includes a market division forecasting unit 8 (see FIG. 1), and is composed of one or more bidding areas. The contract forecast price Pc (area forecast price Pca) may be calculated in consideration of the market division due to the restriction of the amount of electric energy that can flow the inter-regional interconnection line connecting the regions. FIG. 9A is a diagram showing a calculation logic of the area predicted price Pca according to the embodiment of the present invention. When this market division occurs, the contract price (estimated area price) of the divided area is usually higher than that in the other case, and the price is different from the case where the market division does not occur. Therefore, the electric power price prediction system 1 makes a market division prediction.

Specifically, in some embodiments, as shown in FIG. 9A, the past performance database Dr further relatives to the approximately quantitative performance information Qr in the past of the interregional interconnection 96 (see FIG. 9B). The interconnection line usage record information Lr, which is the amount of power used, is associated and stored. Further, the electric power price prediction system 1 further includes an interconnection line allowance acquisition unit 81, an interconnection line usage record acquisition unit 82, a division prediction execution unit 83, and an area prediction price calculation unit 84.

The interconnection line allowance acquisition unit 81 acquires the transmission allowance Lc of the interregional interconnection line 96. In the embodiment shown in FIG. 9A, the interconnection line allowance acquisition unit 81 transmits power of the interregional interconnection line 96 from the interconnection line information database Dl that stores the transmission allowance Lc of the interregional interconnection line 96. The allowable amount Lc is acquired.

The interconnection line usage record acquisition unit 82 includes the buy bid contract quantitative forecast Qbc and the sell bid contract quantitative forecast Qcs calculated by the approx. Quantitative forecast calculation unit 5, and the buy bid contract quantitative forecast stored in the past record database Dr. Based on each comparison with the actual Qcr (b) and the selling bid contract quantitative forecast actual Qcr (s), the interconnection line usage actual information Lr is acquired from the past actual database Dr.
The division prediction execution unit 83 predicts whether or not market division will occur based on the comparison between the transmission allowance Lc of the inter-regional interconnection line 96 and the interconnection line usage actual information Lr.
That is, the interconnection line usage record information Lr acquired by the interconnection line usage record acquisition unit 82 is used as the predicted value of the interconnection line usage amount at the target delivery date and time, and this predicted value is compared with the transmission allowable amount Lc. Based on this, we predict the possibility of market disruption.

For example, when the transmission capacity Lc of the inter-regional interconnection line 96 connecting Hokkaido and Honshu is c1, and the predicted value of the interconnection line usage (interconnection line usage record information Lr) is x1, then c1 <x1. Judge that there is a high possibility that market division will occur. On the contrary, when c1> x1, the possibility of market division is low. The possibility of this market division may be determined stepwise, such that the possibility of division may occur at a medium level or a low level, depending on the magnitude of the difference in c1-x1.

In some embodiments, the output unit 14 (see FIG. 1) sets the result of the prediction by the division prediction execution unit 83 of the interregional interconnection line 96 as shown in FIG. 9B. Output with position information. FIG. 9B is an example of the screen display G showing the information of the inter-regional interconnection line 96 including the possibility of market division according to the embodiment of the present invention mapped to a map. In the embodiment shown in FIG. 9, the output unit 14 outputs the possibility of division (result of the above prediction) and the position information of the inter-regional interconnection line 96 together with the map information, and each inter-regional interconnection. On the screen display G that schematically maps the interconnection line 96 to a map (not shown) such as a map of Japan, the identification information of the interconnection line 96 between regions (such as the HVDC Hokkaido-Honshu interconnection facility) and its division are possible. Gender, the predicted value of the interconnection line usage, and the transmission allowance Lc are displayed in an overlapping manner. In this way, the result of the market division prediction (division possibility) is shown together with the position information of the interregional interconnection line 96, such as showing the result of the market division prediction in the map information showing the position of the interregional interconnection line 96. By outputting to an output device such as a display, it is possible to grasp the risk of market division at a glance. The location information may be the name of the inter-regional interconnection line 96, and the output unit 14 may display a list of the name of the inter-regional interconnection line 96 and the possibility of division.

The area forecast price calculation unit 84 calculates the area forecast price Pca, which is the contract forecast price Pc for each bid area when the market division is predicted to occur (see FIGS. 10 to 11 described later).

According to the above configuration, the contract forecast price Pc for each bidding area when the market division occurs can be calculated, and the contract forecast price Pc can be calculated more accurately.

Next, an embodiment of calculating the contract forecast price Pc and the area forecast price Pca for each of the one-day-ahead market and the current-day market constituting the variable market will be described in detail with reference to FIGS. 10 to 11. FIG. 10 is a diagram showing a calculation logic for calculating a contract forecast price Pc in the one-day-ahead market according to an embodiment of the present invention. Further, FIG. 11 is a diagram showing a calculation logic for calculating the contract forecast price Pc in the market on the day according to the embodiment of the present invention.
In some embodiments, the contract forecast price calculation unit 6 includes a one-day market price forecast unit 6p that calculates the contract forecast price Pc in the one-day market and a same-day market price forecast unit 6 that calculates the contract forecast price Pc in the current market. 6t and the like (see FIG. 1).

In the embodiment shown in FIG. 10, in the past actual record R, the past actual database Dr has the buy bid contract quantitative forecast actual Qcr (b) and the sell bid contract quantitative forecast actual Qcr (s) with respect to the contract forecast price Pc. The market division forecast result by the market division prediction unit 8 at that time (the possibility of market division is large, medium, small, etc.) and the market breakdown prediction result by the market breakdown prediction calculation unit 7. The market breakdown forecast results and the market division occurrence results are stored in association with each other. Then, the one-day-ahead market price forecasting unit 6p includes the buy bid contract quantitative forecast Qcb, the sell bid contract quantitative forecast Qcs, the market division forecast result and the market breakdown forecast result, and the buy bid contract quantitative forecast stored in the past performance database Dr. It is a past actual record R in which the actual Qcr (b), the selling bid contract quantitative forecast actual Qcr (s), the market division forecast actual, and the market breakdown forecast actual match, and there is no market division in the market division forecast actual. The past performance record R is extracted, and the contract forecast price Pc in the one day ago market is calculated based on the contract price performance information in the one day ago market included in the contract price performance information Pr of the extracted past performance record R.

In addition, if the possibility of market division is above a certain level (high or above, medium or above, small or above, etc.) in the market division forecast result, there is a market division due to the market division occurrence record in the target bidding area and target delivery date and time. The past actual record R to be obtained is extracted, and the area predicted price Pca is calculated based on the area price actual of the area included in the contract price actual information Pr of the extracted past actual record R.
In the embodiment shown in FIG. 10, the contract predicted price Pc (without market division) and the area predicted price Pca (with market division) calculated as described above are presented to the user who uses this system. At this time, as shown in FIG. 10, information considered to be useful, such as the market division forecast result and the market breakdown forecast result, may be presented together.

In the embodiment shown in FIG. 11, in the past record R, the past record database Dr has a buy bid contract quantitative forecast result Qcr (b) and a sell bid contract quantitative forecast result Qcr (s) with respect to the contract forecast price Pc. And, the market division forecast result and the contract price result information (or area price result) in the market one day ago are stored in association with each other. Then, the market price forecasting unit 6t on the day includes the buy bid contract quantitative forecast Qccb, the sell bid contract quantitative forecast Qcs, the market division forecast result, the contract forecast price Pc in the market one day before, and the buy stored in the past performance database Dr. Past results in which the bid contract quantitative forecast results Qcr (b), the sell bid contract quantitative forecast results Qcr (s), the market division forecast results, and the contract price performance information Pr (or area price results) in the previous day's market match. The record R is extracted, and the contract forecast price Pc in the market on the day is calculated based on the contract price actual information Pr of the extracted past actual record R.

In the embodiment shown in FIG. 11, the search conditions are the buy bid contract quantitative forecast Qcb, the sell bid contract quantitative forecast Qcs, the market division forecast result, and the contract forecast price Pc (or area forecast price Pca) in the market one day before. The past record database Dr was searched. However, the present invention is not limited to the present embodiment. In some other embodiments, the buy bid contract quantitative forecast Qcc, the sell bid contract quantitative forecast Qcs, and the sell bid contract quantitative forecast Qcs in the market on the day are calculated based on the buy bid contract quantitative forecast Qcb, the sell bid contract quantitative forecast Qcs, and the market division prediction result, respectively. Then, the market bid amount information on the day may be used, and the past performance database Dr may be searched using the market bid amount information on the day and the contract forecast price Pc in the market one day before as search conditions.

Although the embodiment in which the electric power price prediction system 1 calculates the contract predicted price Pc has been described above, the electric power price prediction system 1 further includes a function (bid pattern determination unit 12) for determining a bid pattern. Also good (see Figure 1). That is, in some embodiments, the electric power price prediction system 1 has a bid pattern determination unit 12 that determines a bid pattern (bid policy) for a plurality of markets (same-day market, one-day-ahead market) constituting a variable market. , Further prepare. In the present embodiment, a profit pursuit mode, a supply shortage avoidance mode, and a stable operation mode are prepared, and a bid pattern according to the mode selected by the user of the system is determined and presented to the user. The electric power trading system 9 (see FIG. 1) may be automatically bid according to the determined bidding pattern.

The above-mentioned profit pursuit mode is a mode prepared when it is desired to pursue the sale at the high price desired by the electric power seller 91 or the purchase at the low price desired by the electric power purchaser 92. For example, when the contract forecast price Pc or the area forecast price Pca in the one-day market calculated as described above is higher than the contract forecast price Pc in the current day market, the bid pattern determination unit 12 is sold by the user. If the electric power company 91 gives priority to bidding on the market one day before, while if the user is the electric power purchasing company 92, the bidding on the same day market is prioritized and the bidding pattern is proposed. On the contrary, for example, when the contract forecast price Pc or the area forecast price Pca in the one-day market is lower than the contract forecast price Pc in the current day market, the bid pattern determination unit 12 allows the user to sell electricity. If the user 91, the bid to the market on the day is prioritized, while if the user is the power purchaser 92, the bid to the market one day before is prioritized and the bidding pattern is proposed. At this time, the above comparison may be made in consideration of the trading fee. Further, in order to avoid a decrease in efficiency of the power generation facility due to short-time operation, the power seller 91 estimates the time zone width in which the decrease in efficiency is small from the power source configuration information or obtains it by the user in advance, and obtains the time. It is recommended to bid for the time zone (electric power products during the time of soaring prices) where the maximum demand is in the band width. At this time, a bidding method for collectively bidding a plurality of time-consecutive electric power products, such as a block bidding on the Japan Electric Power Exchange (JEPX), may be used together. On the other hand, the power purchaser 92 that owns the power storage equipment may be recommended to bid in the time zone when the minimum demand is reached (electric power product in the time zone when the price falls). It should be noted that whether or not the power storage equipment is owned is confirmed in advance for each power purchaser 92 by providing a check column when the mode is selected by the user operation described above.

The above-mentioned power shortage avoidance mode is a mode prepared when both the power selling business operator 91 and the power purchasing business operator 92 want to prioritize avoidance of expensive inbound clearing due to supply shortage. For example, level 1 may be set when it is desired to avoid as much as possible, and level 2 may be set when it is absolutely necessary to avoid it. Further level settings may be provided in this mode. Then, for example, at level 1 (avoidance as much as possible), the power seller 91 sells only the amount of power that has sufficient or appropriate power generation surplus capacity in the market one day before, and the surplus power while observing the situation in the market on the day. You may propose to sell the electricity. On the other hand, the electric power purchaser 92 purchases (bids) the electric power corresponding to the above-mentioned buy bid contract quantitative forecast Qcb, which is the result of the electric power demand forecast of the transmission destination to which the electric power is transmitted, in the market one day before, and is insufficient. You may propose to buy (bid) in the market on the day when the minute is generated. Further, at level 2 (absolute avoidance), only the amount of power that has sufficiently or appropriately left the surplus power generation capacity in the market one day ago is sold to the power selling company 91, while the power purchasing company 92 is sold to one. In the day-ahead market, it is possible to purchase (bid) an amount that exceeds the bid bid contract quantitative forecast Qcb by a predetermined amount, and in the day-ahead market, it may be proposed to sell the surplus electricity while observing the situation.

The above stable operation mode is a mode that can be selected by the power sales company 91 (power generation company), and can be operated stably by reducing the number of starts and stops that put a load on the power generation equipment (equipment). This mode is prepared when you want to. Therefore, the amount that can maintain the rated output is continuously executed for 24 hours. At this time, for example, if the power sales company 91 owns a power generation facility with good responsiveness such as a renewable power generation facility that uses a gas engine and a storage battery together, the unit price of power sale is high because of the last-minute transaction. Propose to prioritize trading in the Deaf day market. More specifically, even if the transaction is performed by adjusting the balance between the market on the day and the market one day before, depending on the size of the difference between the buy bid contract quantitative forecast Qcc and the sell bid contract quantitative forecast Qcs (Qcc / Qcs). good. For example, as shown in FIG. 12, when Qcb / Qcs> 1, there is a high possibility that the rated output can be sold out even in the market on the day, so the ratio of the amount of electric power to bid on the market on the day is increased. Correspondingly, we will make a proposal to reduce the ratio of the amount of electricity bid in the market one day ago. In the case of Qcb / Qcs ≤ 1, there is a high possibility that the rated output cannot be sold out in the market on the day, so increase the ratio of the amount of power to bid in the market one day before and bid in the market on the day accordingly. Make proposals that reduce the ratio of electric energy. In addition, if the power sales company 91 owns a power generation facility with poor responsiveness such as BTG, it is easy to make a plan in advance and maintain efficiency even if the unit price of power sales may be reduced one day before. Propose to prioritize trading in the market. Priority is given to block bidding in particular. In addition, the shortfall in the rated output will be supplemented in the market on the day.

According to the above-described configuration, the electric power price prediction system 1 can make an appropriate bid in a variable market.

Further, in some embodiments, the electric power price prediction system 1 may cooperate with the control unit 91c that controls the power generation plant 91p. This embodiment will be described with reference to FIG. FIG. 13 is a diagram showing an electric power price prediction system 1 linked with a power plant 91p according to an embodiment of the present invention.

Specifically, as shown in FIG. 13, the electric power price prediction system 1 is a power generation plant 91p including the actual power generation unit price generated by the power generation plant 91p (the AI-equipped plant in FIG. 13) owned by the electric power seller 91. A bid condition determination unit 13 (bid optimization AI in FIG. 13) for determining bid conditions in consideration of driving performance information is further provided. The power plant 91p of the present embodiment is operated and controlled by an operation optimization AI (control unit 91c) configured to perform operation control so as to achieve optimization such as operation efficiency based on machine learning. Then, the operation optimization AI acquires the fuel price information Fc from the fuel information database Df, and acquires the environmental forecast information Ec of the area where the own plant is located from the environmental information database De, so that the actual operation under the operation control is performed. The operation information including the power generation unit price (for example, yen / kWh) of is calculated.

On the other hand, the bid condition determination unit 13 uses the calculated contracted forecast price Pc in the same-day market and the contracted forecasted price Pc or area forecasted price Pca in the one-day-ahead market as described above to provide unit price information (for example, yen) for selling electricity in each market. / KWh). The buy bid contract quantitative forecast Qcb, the sell bid contract quantitative forecast Qcs, and the market breakdown forecast result are acquired as bid amount information in each market. Further, the bid condition determination unit 13 is connected to the above-mentioned operation optimization AI, and acquires operation information including the power generation unit price from the operation optimization AI.

In this way, the bid condition determination unit 13 has the power sale unit price information and bid amount information in each market obtained as various prediction results already calculated, and the actual power generation unit price according to the plant characteristics controlled by the operation optimization AI. In consideration of the driving information including, for example, the bidding pattern that realizes the most economical and most economical driving is determined. The bid condition determination unit 13 may create and execute an operation plan that prioritizes _{not only economic efficiency but also environmental regulation values such as CO 2 and NOx by selecting a mode by the user.}

In addition, the operation optimization AI acquires the contract result (contract power amount for each power product) in the power trading system 9, creates a power generation amount-time map based on the contract result, and each power generation according to this map. Create an operation plan for the equipment and execute the operation plan. In the case of a power generation company (power sales company 91) that owns a gas turbine and BGT by utilizing various information related to operation such as startability of power generation equipment, fuel type, load change rate, and outside temperature, for example. Since the temperature is low from 6:00 to 9:00, the gas turbine alone generates all the electricity, and the exhaust heat warms up the BTG. From 9:00 to 16:00, the gas turbine is X (kWh) and the BTG is Y (kWh). It will generate electricity in combination, and from 16:00 to 22:00, the boiler fuel will be changed from coal-only operation to co-firing operation of coal and woody biomass.

According to the above configuration, the bidding conditions that are optimal for the operation of the power plant 91p can be determined by determining the bidding conditions in consideration of the operation record information of the power plant 91p.

The present invention is not limited to the above-described embodiment, and includes a modified form of the above-described embodiment and a combination of these embodiments as appropriate.

1 Electricity price forecast system 12 Bid pattern determination unit 13 Bid condition determination unit 14 Output unit 2 Forward contract total amount acquisition unit 3 Environmental forecast acquisition unit 4 Approximate quantitative result acquisition unit 4b Buy bid Quantitative result acquisition unit 4s Sell bid Quantitative result Acquisition unit 44 Fuel price acquisition unit 45 Power generation capacity forecast acquisition unit 46 Power generation capacity calculation unit 47b Buy bid contract quantitative prediction error calculation unit 47s Sell bid contract quantitative prediction error calculation unit 48 Transmission destination information acquisition unit 5 Transmission destination information acquisition unit 5 Approximate quantitative forecast calculation unit 5b Buy Bid Contract Quantitative Forecast Calculation Unit 5s Sell Bid Quantitative Prediction Calculator 6 Contract Prediction Price Calculator 6p One Day Before Market Price Prediction 6t Same-day Market Price Prediction 7 Market Breakdown Forecast Calculation Unit 71 Breakdown forecast calculation unit 73 Selling bid breakdown performance acquisition unit 74 Selling bid market breakdown forecast calculation unit 8 Market division forecasting unit 81 Interconnection line allowance acquisition unit 82 Interconnection line usage actual acquisition unit 83 Division forecast execution unit 84 Area forecast price Calculation unit 9 Power trading system 91 Power seller 91p Power plant 91c Control unit (operation optimization AI)
92 Power purchaser 93 Contractor Consumer 96 Interregional interconnection

Ac Power generation capacity forecast Ar Power generation capacity performance information C Transmission destination information Cr Transmission destination information database Dcb Bid market breakdown forecast Qcr (b) Buy bid contract quantitative forecast result Qcr (s) Sell bid contract quantitative forecast result Qrb Buy Bid Approximate Quantitative Prediction Qrs Sell Bid Approximate Quantitative Prediction Dcs Sell Bid Market Breakdown Forecast De Environmental Information Database Dl Interconnection Information Database Dm Forward Information Database Dp Power Supply Configuration Database Dr Past Results Database Drb Buy Bid Breakdown Results Drs Sell Bid Breakdown Results Ec Environmental forecast information Er Environmental performance information Fc Fuel price information Fr Fuel price performance information ID Business operator Lc Transmission capacity Lr Interconnection line usage performance information Pc Contract price forecast price Pca Area forecast price Pr Contract price performance information Qc Quantitative price forecast Qcc Buy Bid Approximate Quantitative Prediction Qcs Sell Bid Approximate Quantitative Prediction Qr Approximate Quantitative Results Qrb Buy Bid Approximate Quantitative Results Qrs Sell Bid Approximate Quantitative Results Qm Forward Contract Total Amount Qrm Forward Approximate Quantitative Results Information Qrp Achievement information R Past achievement record

Claims (19)

A power price prediction system that calculates the contract forecast price in a variable market where the market price of electricity delivered from the power seller to the power purchaser fluctuates depending on the bidding status on the target delivery date and time, which is the desired delivery date and time. There,
Environmental performance information including past weather information, about quantitative performance information that is the sum of past power traded in each of the variable market and the forward market that trades power before the trading date of the variable market, said. A past record database that stores the contract price record information, which is the record of the contract price of electric power traded in the past in the variable market, by the delivery date and time and by the bid area, and
A forward contract total amount acquisition unit that acquires the forward contract total amount, which is the amount of electric power delivered in the target bid area, which is the desired bid area at the target delivery date and time traded in the forward market.
An environmental forecast acquisition unit that acquires environmental forecast information including weather information of the target delivery date and time in the target bid area, and
Based on the comparison between the environmental forecast information of the target delivery date and time in the target bid area and the environmental record information stored in the past record database, the target delivery date and time in the target bid area can be obtained from the past record database. The Approximate Quantitative Results Acquisition Department that acquires the Approximate Quantitative Results Information corresponding to the time,
By subtracting the total amount of forward contracts from the actual quantitative information, the quantitative forecast calculation unit that calculates the quantitative forecast, which is the predicted value of the electric energy traded in the variable market,
It is provided with a contract forecast price calculation unit that acquires the contract price actual information from the past record database based on the contract quantitative forecast and calculates the contract price forecast price based on the acquired contract price actual information. A featured electricity price forecasting system. The past performance database further stores the contract price performance information in association with the contracted price forecast calculation unit, which is the contracted quantitative forecast calculated in the past, or the contracted price forecast information. And
The contract price forecast calculation unit obtains the contract price actual information from the past actual database based on the comparison between the contract quantitative forecast and the contract quantitative forecast actual or the contract quantitative actual information stored in the past actual database. The electric power price prediction system according to claim 1, wherein the acquisition is performed. The fuel price acquisition department that acquires fuel price information of the fuel used for power generation by the power seller,
Further equipped with a power generation capacity prediction acquisition unit for acquiring a power generation capacity prediction at the target delivery date and time of the power plant owned by the power sales business operator.
The past record database is composed of the environmental record information, the contracted quantitative record information, the contract price record information, the power generation capacity record information for each past date and time in the power seller, and the fuel price information in the past. Along with memorizing a certain fuel price actual information by associating it with each other
The said about quantitative result acquisition department
Based on the comparison between the environmental forecast information and the environmental performance information, the buy bid contracted quantitative result acquisition unit that acquires the contracted quantitative result information from the past record database and uses the bid bid contracted quantitative result as the buy bid contracted quantitative result.
In addition to the comparison between the environmental forecast information and the environmental actual information, the past is based on the comparison between the power generation capacity forecast and the power generation capacity actual information and the comparison between the fuel price information and the fuel price actual information. It has a selling bid contract quantitative result acquisition unit that acquires the above-mentioned approximate quantitative record information from the record database and uses it as a sell bid contract quantitative record.
The approximate quantitative prediction calculation unit
The Bid Bid Quantitative Prediction Calculator, which calculates the Bid Quantitative Prediction by subtracting the total amount of the forward contracts from the Bid Quantitative Results.
The electric power price forecast according to claim 2, further comprising a sell bid contract quantitative forecast calculation unit that calculates a sell bid contract quantitative forecast by subtracting the forward contract total amount from the sell bid contract quantitative actual information. system. A power source configuration database that stores power source configuration information, which is information on equipment constituting the power plant owned by the power seller, and stop information of the power plant for each power seller, and a power source configuration database.
A power generation capacity calculation unit that calculates the power generation capacity of the power sales business operator based on the power generation configuration information and the stop information of the power plant is further provided.
The power price prediction system according to claim 3, wherein the power generation capacity prediction acquisition unit acquires the power generation capacity calculated by the power generation capacity calculation unit as the power generation capacity prediction. In the past, the past performance database used the buy bid contract quantitative forecast result, which is the bid bid contract quantitative forecast calculated based on the bid bid contract quantitative forecast, to calculate the bid bid contract quantitative forecast result. It is stored in association with the said quantitative result information, which is the purchased bid contract quantitative result.
Based on the comparison between the buy bid contract quantitative forecast and the buy bid contract quantitative forecast result stored in the past record database, the buy bid contract quantitative forecast result and the contract quantitative result information are obtained from the past record database. In addition to being acquired, it is further equipped with a buy bid contract quantitative prediction error calculation unit that calculates a buy bid contract quantitative forecast error based on the acquired bid bid contract quantitative forecast result and the contract quantitative result information.
The electric power price prediction system according to claim 3 or 4, wherein the buy bid contract quantitative prediction calculation unit calculates the buy bid contract quantitative forecast using the buy bid contract quantitative prediction error. In the past, the past performance database uses the sell bid contract quantitative forecast result, which is the sell bid contract quantitative forecast calculated based on the sell bid contract quantitative forecast, to calculate the sell bid contract quantitative forecast result. It is stored in association with the above-mentioned about quantitative result information, which is the sell bid about quantitative result.
Based on the comparison between the sell bid contract quantitative forecast and the sell bid contract quantitative forecast result stored in the past record database, the sell bid contract quantitative forecast result and the contract quantitative result information are obtained from the past record database. In addition to being acquired, it is further equipped with a sell bid contract quantitative prediction error calculation unit that calculates a sell bid contract quantitative forecast error based on the acquired sell bid contract quantitative forecast result and the contract quantitative result information.
The electric power price according to any one of claims 3 to 5, wherein the selling bid contract quantitative prediction calculation unit calculates the selling bid contract quantitative prediction using the selling bid contract quantitative prediction error. Prediction system. A power transmission destination information database for storing power transmission destination information including the location of the power transmission destination regarding the power transmission destination to which the power purchaser transmits the power, and a power transmission destination information database.
Further, a power transmission destination information acquisition unit for acquiring the power transmission destination information of a plurality of the power transmission destinations located in the target bidding area from the power transmission destination information database is further provided.
The environmental forecast acquisition unit acquires the environmental forecast information for each of the locations of the plurality of transmission destinations, and obtains the environmental forecast information.
The purchase bid contract quantitative prediction calculation unit is based on the comparison between each of the environmental forecast information for each location of the transmission destination and the environmental record information stored in the past record database, and from the past record database. The electric power price prediction system according to claim 3, wherein the total of the acquired quantitative actual information is used as the buy bid contract quantitative forecast. The variable market includes the same-day market that determines the market price on the delivery date and the one-day market that determines the market price one day before the delivery date.
The past record database contains the environmental record information, the buy bid contract quantitative forecast record, the sell bid contract quantitative forecast record, and the current day contract quantitative record which is the electric energy for each past date and time traded in the market on the day. The information and the one-day-ahead quantitative actual information, which is the amount of electricity for each past date and time traded in the one-day-ahead market, are stored in association with each other.
The electricity price forecasting system further
Based on the comparison between the bid bid contract quantitative forecast and the environmental forecast information, and the bid bid contract quantitative forecast result and the environmental record information stored in the past record database, the same-day contract quantitative result information and the above-mentioned one. The Bid Breakdown Results Acquisition Department, which acquires information on the quantitative results of the day before,
Based on the current-day contract quantitative performance information and the one-day-previous quantitative performance information acquired by the bid breakdown performance acquisition unit, the above-mentioned traded in each of the same-day market and the one-day-ahead market constituting the variable market. In addition to calculating the purchase bid breakdown results, which is the breakdown of the amount of electricity, the buy bid market breakdown forecast, which is the breakdown of the same-day market and the one-day-ahead market in the purchase bid contract quantitative forecast, is calculated based on the purchase bid breakdown results. Bid market breakdown forecast calculation department and
Based on the comparison between the sold bid contract quantitative forecast and the environmental forecast information, and the sold bid contract quantitative forecast result and the environmental record information stored in the past record database, the same-day contract quantitative result information and the above-mentioned one. The variable market is constructed based on the sales bid breakdown performance acquisition unit that acquires the day-ahead quantitative performance information, the current-day contract quantitative performance information acquired by the sales bid breakdown performance acquisition department, and the one-day contract quantitative performance information. The current bid breakdown, which is a breakdown of the amount of electricity traded in each of the same-day market and the one-day-ahead market, is calculated, and the same-day market in the quantitative forecast of bid contracts is calculated based on the bid breakdown record information. And a bid bid market breakdown forecast calculation unit that calculates the bid bid market breakdown forecast, which is the breakdown of the previous day's market.
The contract forecast price calculation unit is based on the buy bid contract quantitative forecast, the bid market breakdown forecast, the sell bid contract quantitative forecast, and the sell bid market breakdown forecast, and at least one of the same-day market and the one-day-ahead market. The electric power price prediction system according to any one of claims 3 to 7, wherein the contract forecast price is calculated. The past performance database further indicates the past power consumption of the inter-regional interconnection line connecting the regions composed of one or more of the bidding areas with respect to the approximately quantitative performance information. The information is associated and memorized,
The interconnection line allowance acquisition unit that acquires the transmission allowance of the inter-regional interconnection line, and the interconnection line allowance acquisition unit.
Based on the comparison between the buy bid contract quantitative forecast and the sell bid contract quantitative forecast, and the buy bid contract quantitative forecast result and the sell bid contract quantitative forecast result stored in the past performance database, the past The interconnection line usage record acquisition unit that acquires the interconnection line usage record information from the track record database, and
A division prediction execution unit that predicts whether or not a market division will occur based on a comparison between the transmission capacity of the inter-regional interconnection line and the actual information on the usage amount of the interconnection line.
The electric power according to claim 8, further comprising an area forecast price calculation unit that calculates an area forecast price, which is the contract forecast price for each bid area when the market division is predicted to occur. Price forecasting system. The electric power price prediction system according to claim 9, further comprising an output unit that outputs the result of the prediction by the division prediction execution unit and the position information of the inter-regional interconnection line. The electric power price prediction system according to any one of claims 1 to 10, further comprising a bidding pattern determining unit for determining a bidding pattern for bidding on a plurality of markets constituting the variable market. Claims 1 to 1, further comprising a bid condition determination unit that determines bid conditions in consideration of the operation record information of the power plant including the power generation unit price record generated by the power plant owned by the power seller. The electric power price prediction system according to any one of 11. The interconnection line usage record acquisition unit includes the buy bid contract quantitative forecast and the sell bid contract quantitative forecast, and the buy bid contract quantitative forecast result and the sell bid contract quantitative forecast result stored in the past record database. 9. The similarity with the above is obtained for each record managed in the past record database, and the interconnection line usage record information is acquired from the past record database based on the similarity. The power price forecasting system described in. The purchase bid breakdown result acquisition unit determines the degree of similarity between the purchase bid contract quantitative forecast and the environmental forecast information and the buy bid contract quantitative forecast record and the environmental record information stored in the past record database. Obtained for each record managed in the performance database, and based on the similarity, the current day's quantitative performance information and the one day's previous quantitative performance information are acquired from the past performance database.
The sell / bid breakdown record acquisition unit records the degree of similarity between the sell / bid contract quantitative forecast and the environmental forecast information and the sell / bid contract quantitative forecast result and the environmental record information stored in the past record database. The power price prediction system according to claim 8, wherein the power price prediction system is obtained for each time, and based on the similarity, the current day's quantitative actual information and the one day's previous quantitative actual information are acquired from the past actual database. In the contract forecast price calculation, the degree of similarity between the contract quantitative forecast and the contract quantitative forecast result or the contract quantitative result information stored in the past record database is obtained for each record managed in the past record database. The power price prediction system according to any one of claims 2 to 14, wherein the contract price actual information is acquired from the past actual database based on the similarity. The approximately quantitative result acquisition unit determines the degree of similarity between the acquisition condition including the environmental forecast information of the target delivery date and time in the target bid area and the information including the environmental record information stored in the past record database. It is characterized in that it is obtained for each record managed in the past performance database, and based on the similarity, the approximate quantitative performance information corresponding to the time of the target delivery date and time in the target bidding area is acquired from the past performance database. The power price prediction system according to any one of claims 1 to 15. The buy bid breakdown record acquisition unit and the sell bid breakdown record acquisition unit obtain the same-day contracted quantitative record information and the one-day-a-day previous quantitative record information included in the record exceeding the preset threshold of similarity. Obtained from the past record database,
The buy / bid market breakdown forecast calculation unit and the sell / bid market breakdown forecast calculation unit are a set of a plurality of acquired same-day contractive actual information when a plurality of the records have the same degree of similarity exceeding the threshold value. This is a value calculated by performing a predetermined calculation for each of the current day's approximately quantitative result set and the one-day's approximately quantitative result set, which is a set of a plurality of acquired one-day's approximately quantitative results information. A claim characterized in that the buy bid breakdown result or the sell bid breakdown result is calculated based on the current day's approximate quantitative representative value based on the set and the one day's approximate quantitative result representative value based on the one day ago quantitative result set. 14. The power price prediction system according to 14. The contract forecast price calculation unit acquires the contract price actual information included in the record exceeding the preset threshold value of the similarity from the past actual database, and also obtains the contract price actual information.
When a plurality of the records have the similarity exceeding the threshold value, a predetermined calculation is performed on the contract price actual set which is a set of the contract price actual information to be acquired, and the contract price actual representative value is calculated. The electric power price prediction system according to claim 15, further comprising calculating the contract forecast price based on the contract price actual representative value. The approximate quantitative achievement acquisition unit acquires the approximate quantitative achievement information included in the record exceeding the preset threshold value of the similarity from the past achievement database.
When a plurality of the records have the same degree of similarity exceeding the threshold value, the representative value of the approximate quantitative result calculated by performing a predetermined calculation on the approximately quantitative actual set which is a set of the plurality of acquired approximately quantitative actual information is obtained. The electric power price prediction system according to claim 16, wherein the information is acquired as the approximately quantitative actual information.

Images