JP7701198B2 - Power selling control device, power selling control system, and power selling control method - Google Patents

Description

The present invention relates to a bidding process for the sale of electricity from a power generation site, including variable power sources such as renewable energy, and a power sales control device, power sales control system, and power sales control method for controlling the power generation site in accordance with the bidding agreement.

In recent years, variable power sources, such as solar power generation facilities, whose power output fluctuates due to external factors, are increasingly being used as one of the energy sources in energy consumption activities in homes and other areas. In these cases, it is desirable to minimize the expected value of energy costs, taking into account the effects of changes in the output of variable power sources and the diversity of load change patterns.

Patent document 1 describes a method in which the value of an optimal control operation sequence is calculated by simulation based on multiple weighted load pattern prediction information, and the next operation that minimizes the evaluation value for a specified period is set as the control command.

Patent Document 2 also describes a technology that determines the start/stop and output allocation of load-controllable power generation equipment by simultaneously considering everything from energy (kWh) supply to frequency adjustment, taking into account the uncertainty of such loads and renewable energy generation power for generators, variable output renewable energy power generation equipment that can be suppressed, and loads that can be controlled.

Patent No. 5813544 Special table number 2019-535225

In renewable energy power generation businesses, the amount of electricity to be delivered during the future electricity delivery period must be agreed upon in the market or bilaterally by a specified time. The future electricity delivery period is specified by a start time and duration, for example, 30 minutes from 13:00. The specified time could be a specified time on the day before the electricity delivery period, or one hour before the start of the electricity delivery period.

Meanwhile, during the actual delivery period, penalties are imposed according to the discrepancy between the contracted amount and the actual supply amount. Therefore, in order to eliminate the discrepancy between the actual amount of power generated by renewable energy power generation facilities, which do not always generate power as predicted, and the contracted amount, it is necessary to activate the procurement adjustment capabilities of storage batteries, private power generation facilities, demand response, etc.

Patent Document 1 shows a method for providing optimal control of storage batteries and other devices that minimizes electricity costs for loads that are expected to stochastically deviate from forecasts, but the method does not assume market transactions and cannot be applied to renewable energy power generation businesses, which require decisions to be made at multiple timings for the same delivery period, such as decisions about the next day or one hour in advance.

Patent Document 2 also shows a method for determining the start/stop and load allocation of controllable power generation equipment that minimizes costs while maintaining the supply/demand balance and system frequency within a specified range. However, the control of power generation equipment is based on the deviation from the target frequency that occurs in response to an imbalance in the supply/demand balance, and is premised on control that increases/decreases output to eliminate the frequency deviation. In contrast, the error between the contracted amount and the amount of power generated by renewable energy power generation companies is unrelated to fluctuations in the system frequency, so the method disclosed in Patent Document 2 cannot be applied to control of storage, etc. to avoid errors between the amount of power generated and the contracted amount.

The present invention aims to bid so that the electricity generated by a power generation site can be sold in an optimal manner, and to control the power generation site according to the contracted amount of the bid.

In order to solve the above-mentioned problems, the power selling control device of the present invention is configured to include a neural network that performs a Fourier transform on power generation output information of a variable power source, which is a solar power generation facility , converts the power generation output information into intensity information and phase information in a frequency space, and machine-learns the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period. The power selling control device is characterized in that it is equipped with: a power generation prediction means that generates a power generation amount prediction time series that predicts the power generation amount of a resource for a predetermined period using the neural network and an inverse Fourier transform at the bidding stage for the day-ahead market or the same-day market; a bidding planning means that determines a transfer amount that can be provided to the power system from the power generation amount prediction time series and submits a bid; a resource control series planning means that determines a control command series for each of the resources that minimizes the error between the contract amount and the actual transfer amount by each of the resources, based on the power generation amount prediction time series, the contract amount determined from the result of the bidding by the bidding planning means, and the state of each of the resources; and a command means that commands the resources to a portion of the control command series determined by the resource control series planning means that can be reflected in the control of each of the resources.

The power selling control system of the present invention is characterized in that it is configured to include a neural network that performs Fourier transform on power generation output information of a variable power source, which is a solar power generation facility , converts the information into intensity information and phase information in a frequency space, and machine-learns the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period, and is equipped with: a power generation prediction means that generates a power generation amount prediction time series that predicts the power generation amount of a resource for a predetermined period using the neural network and an inverse Fourier transform at the bidding stage for the day-ahead market or the same-day market; a bidding planning means that determines a transfer amount that can be provided to the power system from the power generation amount prediction time series and submits a bid; a resource control series planning means that determines a control command series for each of the resources that minimizes the error between the contract amount and the actual transfer amount by each of the resources, based on the power generation amount prediction time series, the contract amount determined from the result of the bidding by the bidding planning means, and the state of each of the resources; and a command means that commands the resources to a portion of the control command series determined by the resource control series planning means that can be reflected in the control of each of the resources.

The power selling control method of the present invention is characterized by comprising the steps of: a power generation prediction means performing a Fourier transform on power generation output information of a variable power source, which is a solar power generation facility , converting the information into intensity information and phase information in a frequency space, and using a neural network that has been machine-learned to understand the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period, and generating a power generation amount prediction time series that predicts the power generation amount of a resource for a predetermined period at the bidding stage for the day-ahead market or the same-day market, using an inverse Fourier transform; a bidding planning means determining a transfer amount that can be provided to the power system from the power generation prediction time series and bidding; a resource control sequence planning means determining a control command sequence for each of the resources that minimizes an error between the contract amount and the actual transfer amount by each of the resources, based on the power generation prediction time series, the contract amount determined from the bidding results by the bidding planning means, and the state of each of the resources; and a command means commanding the resources a portion of the control command sequence determined by the resource control sequence planning means that can be reflected in the control of each of the resources.
Other means will be described in the description of the embodiment of the invention.

According to the present invention, it is possible to bid for electricity generated at a power generation site so that the electricity can be sold appropriately, and to control the power generation site according to the contracted amount of the bid.

1 is a block diagram showing a configuration of a power sale control system and a power sale control device according to an embodiment of the present invention; FIG. 13 is a block diagram showing the configuration of a power sale control system and a power sale control device according to a modified example. 1 is a time chart showing possible values of a planning variable at a target time. 1 is a diagram illustrating a variable power supply and learning by a neural network according to an embodiment of the present invention. FIG.

The following describes in detail the form for implementing the present invention with reference to the drawings. The electricity selling control device of the present invention determines the bid amount for one delivery period as described above, which requires decision-making at multiple timings, such as the day before the start of delivery, one hour before the start of delivery, or during the delivery period, and controls resources such as storage batteries and private power generation facilities to eliminate errors in the actual delivery amount relative to the contract amount determined in advance.

FIG. 1 is a block diagram showing the configuration of a power sale control system 10 and a power sale control device 100 according to this embodiment.
The power sales control system 10 includes a power sales control device 100 that is communicably connected to power generation sites 150 and 150a. The power sales control system 10 sells power generated by a variable power source 151 installed at the power generation site 150.

The variable power source 151 supplies power to the power grid 180, and is, for example, a renewable energy power source such as a solar power generation facility, a wind power generation facility, or a tidal power generation facility. The controllable power source 152 is, for example, a private power generation facility that supplies power to the power grid 180, or a storage battery that charges and discharges the power of the power grid 180. The variable power source 151 and the controllable power source 152 function as resources for supplying power to the power grid 180.

The day-ahead market system 120 and the intraday market system 130 are market systems for procuring electricity to be supplied to the power grid 180. The day-ahead market system 120 settles bids made by the power sales control device 100, the demand response business operator 141, and the power generation business operator 142 on the day before the planned target date.

To procure electricity to be supplied to the power grid 180, the intraday market system 130 enters into the bids submitted by the power sales control device 100, the demand response business operator 141, and the power generation business operator 142 by a specified time on the planned target day.

The electricity procured by the day-ahead market system 120 and the intraday market system 130 is purchased from bids made by the power selling control device 100, the demand response business operator 141, and the power generation business operator 142. Note that in the figure, the demand response business operator 141 is abbreviated to "DR business operator."

The power generation site 150 and load equipment (not shown) operated by the demand response operator 141, and the power generation equipment (not shown) operated by the power generation operator 142 are connected to a single power grid 180.

The power selling control device 100 includes a day-ahead power generation prediction means 101, a day-ahead bid planning means 103, a power generation time-ahead prediction means 105, a day-ahead bid planning means 107, a high-frequency short-term power generation prediction means 109, an operation planning means 111, an optimization engine 170, and command means 165, 166. The power selling control device 100 is communicatively connected to a variable power source 151 installed at the power generation site 150 and a controllable power source 152 installed at the power generation site 150a. As a result, the power selling control device 100 receives the output 116 of the variable power source 151 and the output 118 of the controllable power source 152. The power selling control device 100 further controls the variable power source 151 by the command value 112, and controls the controllable power source 152 by the command value 113.

The power selling control device 100 bids on the day-ahead market system 120 and the intraday market system 130 to sell electricity generated by the variable power source 151 installed at the power generation site 150 and the controllable power source 152 installed at the power generation site 150a. The power selling control device 100 further plans control commands to the variable power source 151 and the controllable power source 152 based on the results of the market transactions, and issues commands to the variable power source 151 and the controllable power source 152.

When bidding on the day-ahead market system 120, the day-ahead power generation prediction means 101 generates at least one day-ahead power generation forecast series 102 and provides this day-ahead power generation forecast series 102 to the day-ahead bidding planning means 103. This process is performed at a timing when bidding on the day-ahead market system 120 is possible. The day-ahead power generation prediction means 101 functions as a power generation prediction means that generates a power generation amount forecast time series that predicts the power generation amount of a resource for a specified period. The resources include any of storage batteries and private power generation facilities, adjustment power by demand response, and output suppression of the variable power source.

The day-ahead bidding planning means 103 determines the bid amount 104 as the amount of power that can be transferred in each of the day-ahead power generation forecast series 102, assuming different facility operations (time series for output suppression of the variable power source 151 and time series for the output of the controllable power source 152) for each of the day-ahead power generation forecast series 102, and submits a bid to the day-ahead market system 120. When determining the bid amount 104, the day-ahead bidding planning means 103 uses the estimated power price for each power market product (frame) on the planned target date determined by a market price prediction means or estimated market price setting means (not shown), the estimated penalty price for each frame determined by a penalty price prediction means or estimated penalty price setting means (not shown), and the operating cost of the controllable power source 152 (kWh unit price, start/stop cost, depreciation cost, etc.). In other words, the day-ahead bidding planning means 103 is a bidding planning means that determines the amount of transfer that can be provided to the power system 180 from the estimated power generation amount time series of multiple resources at the bidding stage to the day-ahead time market.

The day-ahead market system 120 determines the contract result 114 as a result of the bidding made on the previous day by the day-ahead bid planning means 103 and notifies the same to the intraday bid planning means 107 .
The power generation time-ahead prediction means 105 predicts one or more current day power generation forecast series 106. The power generation time-ahead prediction means 105 functions as a power generation prediction means for generating a power generation forecast time series that predicts the power generation amount of a resource for a predetermined period.

The intraday bid planning means 107 uses the day-ahead market contract results 114 and one or more intraday power generation forecast series 106 predicted by the power generation time-ahead forecasting means 105 to plan the additional bid amount for delivery or the amount of power to be purchased from the intraday market, determine the intraday buying and selling bid amount 108, and submit a bid to the intraday market system 130. At this time, although not shown, similar to the day-ahead bid planning means 103, the estimated transaction price for each intraday market product (frame) on the planned target date determined by the intraday market price prediction means or the estimated intraday market price setting means, and the estimated penalty price for each frame determined by the penalty price prediction means or the estimated penalty price setting means (not shown) are used. The intraday bid planning means 107 is a bid planning means that determines the delivery amount that can be provided to the power system from the power generation amount forecast time series of multiple resources at the bidding stage to the intraday time-ahead market.

The day-ahead bid planning means 103 and the same-day bid planning means 107 also use the output 116 of the variable power source 151 measured by the meter 153, the state 117 of the variable power source 151, the output 118 of the controllable power source 152 measured by the meter 154, and the state 119 of the controllable power source 152. The day-ahead bid planning means 103 and the same-day bid planning means 107 determine the transfer amount that can be commonly provided on the premise of the operation of different resources, such as the controllable power source 152 whose power generation amount can be controlled and the variable power source 151 whose power generation amount fluctuates due to external factors.

The high-frequency short-term power generation forecasting means 109 functions as a power generation forecasting means that generates a high-frequency short-term power generation forecast time series 110 and generates a power generation forecast time series that predicts the power generation amount of a resource for a specified period.

The operation planning means 111 generates a control command series for the variable power source 151 and the controllable power source 152 using the agreement result 115 for the current day's buying and selling bid amount 108 obtained by a specified timing, the agreement result 114 for the previous day's bid amount 104, and the power generation forecast time series 110 of the power generation high frequency short-term forecasting means 109. The operation planning means 111 generates the control command series using the optimization engine 170 and various formulas described below. In other words, the operation planning means 111 functions as a resource control series planning means that determines a control command series for each resource that minimizes the error between the contract amount and the actual delivery amount by each resource, based on the power generation forecast time series, the contract amount determined from the bidding results, and the state of each resource.

The optimization engine 170 is also used by the day-ahead bidding planning means 103 to determine the bid volume 104 and by the day-bidding planning means 107 to determine the day-to-day buying and selling bid volume 108.

The command means 165 commands the variable power source 151 to control the variable power source 151 with the control command series for the variable power source 151 generated by the operation planning means 111. The command means 166 commands the controllable power source 152 to control the controllable power source 152 with the control command series for the controllable power source 152 generated by the operation planning means 111. This allows the power sale control device 100 to sell the power generation output of the power generation sites 150 and 150a in the most optimal way. The command means 165 and 166 function as command means that command the resources to control the parts of the control command series determined by the operation planning means 111 that each resource can reflect in its control.

FIG. 2 is a block diagram showing a configuration of a power selling control system 10A according to a modified example.
The power selling control system 10A includes a power selling control device 100A that is communicatively connected to power generating sites 150 and 150a. A command device 155 is installed at the power generating site 150, and commands the variable power source 151 to generate a control command sequence for the variable power source 151 by an operation planning means 111 of the power selling control device 100A.
A command device 156 is installed at the power generation site 150a, and issues to the controllable power source 152 a control command sequence for the controllable power source 152 generated by the operation planning means 111 of the power sale control device 100A.

Furthermore, the power selling control system 10A is not limited to this configuration, and each means may be realized by multiple computers, and is not limited thereto.

Returning to FIG. 1, an example of a method for implementing the day-ahead bid planning means 103 will be described. Here, the day-ahead bid planning means 103 uses multiple day-ahead power generation forecast series 102 output by the day-ahead power generation forecast means 101 as main inputs, and calculates the bid amount 104 for the electricity sold the next day, taking into account the operation of controllable power sources 152 such as private power generation facilities and storage batteries. One such calculation method is to describe it as a profit optimization problem, and determine the bid amount 104 so as to maximize this.

Equations (101) to (135) are examples of the objective function and constraint equations of such an optimization problem. In this embodiment, a storage battery and a private power generation facility are assumed as the controllable power source 152. These will be described below.

Equation (101) is an example of an objective function for a day-ahead plan, where the optimization variables are the following five sets:

Previous day's bid volume: bidsDA [t]
Power suppression amount of variable power supply: curtailDA[sid,t]
Battery charge amount: actEsCfDA [sid, t]
Discharge amount of storage battery: actEsDfDA [sid, t]
Power output of private power generation equipment: actLclFDA [t]

sid is an identification ID of each forecast series of the day-ahead power generation forecast series 102 output by the day-ahead power generation forecasting means 101. Here, the number of forecast series is expressed by scnoNum. t is identification information of a frame to be traded in the market. The variables constituting the objective function are as follows.

Daily sales: earningsDA[sid,t]
Premium: premiumDA[sid,t]
Operating cost of self-power generation equipment: lflexXpnsDA[sid,t]
Penalty: penaltyDA[sid,t]

The target of optimization is the average value of the daily profit calculated from these terms (the sum of the power sales and premium minus the cost and penalty) for all frames over the entire day-ahead-of-power-generation forecast series 102. Here, the premium refers to a power selling price that is higher than the normal power selling price due to a power shortage or the like.

Equation (102) is a constraint equation for daily sales earningsDA[sid,t]. The variables related to equation (102) are as follows.

Daily sales: earningsDA[sid,t]
Predicted market price: mktPrcOfDA[t]
Delivery amount at time tt: actualGenDA[sid, tt]

Equation (102) shows that the amount of electricity delivered in frame t (the sum calculation part on the right-hand side) multiplied by the predicted market price mktPrcOfDA[t] is the daily revenue from electricity sales in that frame t, earningsDA[sid,t].

The right-hand side of equation (102) is the sum of the actualGenDA[sid, tt] transfer amounts from time tt = t to t + 30, but in the case of constraints such as equation (106) described below, this essentially becomes a single variable, so there is no need to calculate the sum.

Equation (106), which will be described later, is a constraint equation regarding the delivery amount actualGenDA[sid,t] in frame t, and is the sum of actTtlGenDA[sid,tt] for 30 minutes, which is the same as the length of the frame. However, if this is to be managed in more detail, such as for 15 minutes, the sum calculation in equation (102) becomes meaningful.

Equation (103) is a constraint equation for the penalty DA[sid, t]. The variables in equation (103) are as follows.

Penalty: penaltyDA[sid,t]
Excess penalty at frame t: penaltyPlsDA[sid, tt]
Shortage penalty at piece t: penaltyMnsDA[sid, tt]

Equation (103) means that the penalty DA[sid,t] is the sum of the excess penalty and the deficiency penalty in frame t. Like the delivery amount, the penalty DA[sid,t] is defined so that it can be managed more precisely than 30 minutes of the frame length, but the sum calculation in equation (103) has no meaning in the formula described in this embodiment.

Equation (104) is a constraint on the premium premiumDA[sid,t] for frame t. The variables in equation (104) are as follows:

Premium: premiumDA[sid,t]
Predicted or expected value of premium unit price: prmmOfDay[t]
The value of the power generation day-ahead forecast series at time tt: reGenDA[sid, tt]
Battery charging efficiency: esEffc
Discharge amount of the storage battery at time tt: actEsDfDA [sid, tt]
Battery charging efficiency: esEffc
Battery charge amount: actEsCfDA [sid, t]
Power suppression amount of variable power supply: curtailDA[sid,t]

The right-hand side of equation (104) is the predicted or estimated value of the premium unit price prmmOfDay[t] multiplied by the transfer amount from the variable power source 151 (renewable energy source).

In this embodiment, the value of the day-before-power-generation forecast series 102 at time tt is reGenDA[sid, tt]. Although the storage battery can also store power from a private power generation facility, assuming a case where such operation is not performed, the second term on the right side adds the discharge amount actEsDfDA[sid, tt] of the storage battery at time tt.

In addition, to exclude charging and suppression from the premium, the fourth term, the power suppression amount of the variable power source, curtailDA[sid,t], is subtracted from the third term on the right-hand side, the charging amount of the storage battery, actEsCfDA[sid,t].

Equation (105) is a constraint on the operating cost lflexXpnsDA[sid,t] of the private power generation facility in frame t. The variables related to equation (105) are as follows.

Operating cost of self-power generation equipment: lflexXpnsDA[sid,t]
Operating cost of private power generation equipment: localFexPrice
Power output of the private power generating equipment within the frame tt: actLFlex [tt]

The operation unit price of the private power generation facility localFexPrice is assumed to be independent of time and is a constant without a time index, but time may be taken into consideration. In the summation calculation, the sum of the power generation output actLFlex[tt] of the private power generation facility within the frame tt is calculated, converted to the amount of power, and then multiplied by the unit price.

The formula (106) is a constraint formula for the delivery amount. The variables related to the formula (106) are as follows.

Delivery amount at frame t: actualGenDA[sid, tt]
Amount transferred at time tt: actTtlGenDA[sid, tt]

The right side of equation (106) is the sum of actTtlGenDA[sid, tt], which is the amount of power transferred at time tt between time t and t + 30, multiplied by a coefficient to convert it into the amount of power. In this embodiment, the unit of management is from t to t + 30, but it may be managed more precisely, such as t + 5 or t + 15.

Equations (107) to (112) are constraint equations regarding the shortfall in the transfer amount.
Equation (107) defines a delivery amount shortage error genMnusDA[sid, t] that takes a positive value when the delivery amount is insufficient. The variables related to equation (107) are as follows.

Insufficient delivery error: genMnusDA [sid, t]
Bid amount at frame t: reBidsDA[t]
Delivery amount deadband width: pnlteyDeadband
Delivery amount at frame t: actualGenDA[sid, tt]

Expression (107) is the difference between the value that is less than the bid amount reBidsDA[t] at frame t by the transfer amount dead band width pnlt yDeadband and the transfer amount actualGenDA[sid, tt] at time tt.

Equations (108) and (109) are constraint equations that constrain the relationship between the delivery amount deficiency indicator reUdGnDA[sid,t] and the delivery amount deficiency error genMnusDA[sid,t], which is 1 when the delivery amount deficiency error genMnusDA[sid,t] is positive and 0 otherwise. The variables related to equations (108) and (109) are as follows.

Insufficient delivery error: genMnusDA [sid, t]
Predetermined value: bigM
Delivery amount shortage indicator: reUdGnDA[sid, t]

When the delivery amount deficiency error genMnusDA[sid,t] is positive, formula (108) holds whether the delivery amount deficiency indicator reUdGnDA[sid,t] is 1 or 0. However, formula (109) shows that the delivery amount deficiency indicator reUdGnDA[sid,t] is restricted to 1.
On the other hand, when the delivery amount insufficiency error genMnusDA[sid,t] is negative, formula (109) holds regardless of the value of the delivery amount insufficiency indicator reUdGnDA[sid,t]. However, formula (108) restricts the delivery amount insufficiency indicator reUdGnDA[sid,t] to 0. With this definition, when the delivery amount insufficiency error genMnusDA[sid,t] is positive, the delivery amount insufficiency indicator reUdGnDA[sid,t] is restricted to 1, and when it is negative, the delivery amount insufficiency indicator reUdGnDA[sid,t] is restricted to 0.

Equations (110) to (112) are constraint equations regarding the deficiency imbalance amount minusGenDA[sid,t], which takes the same value as the delivery amount deficiency error genMnusDA[sid,t] when the delivery amount deficiency indicator reUdGnDA[sid,t] is 1, and takes a value equal to or greater than zero so as to become 0 when the delivery amount deficiency indicator reUdGnDA[sid,t] is 0. The variables related to equations (110) to (112) are as follows.

Insufficient delivery error: genMnusDA [sid, t]
Deficient imbalance amount: minusGenDA[sid, t]
Predetermined value: bigM
Delivery amount shortage indicator: reUdGnDA[sid, t]

From equations (110) and (111), when the delivery amount shortage indicator reUdGnDA[sid,t] is 1, the delivery amount shortage error genMnusDA[sid,t] and the shortage imbalance amount minusGenDA[sid,t] are constrained to be equal values, and when it is 0, no constraint applies. On the other hand, from equation (112), when the delivery amount shortage indicator reUdGnDA[sid,t] is 0, the shortage imbalance amount minusGenDA[sid,t] is constrained to 0.

In equations (107) and (113), the bid volume reBidsDA[t] for frame t is used. However, if the period is shorter than the length of the frame, this value should be multiplied by a coefficient that takes into account the period length. For example, if the period is 15 minutes, multiply by 15/30.

Equations (113) to (118) are constraint equations that constrain the relationship between the excess delivery error genPlusDA[sid,t], the excess delivery indicator reOvGnDA[sid,t], and the excess imbalance amount plusGenDA[sid,t] in the same manner as equations (107) to (112).

Equation (113) defines the excess delivery amount error genPlusDA[sid, t] that takes a positive value when the delivery amount is insufficient. The variables related to equation (113) are as follows.
Excessive delivery error: genPlusDA[sid,t]
Bid amount at frame t: reBidsDA[t]
Delivery amount deadband width: pnlteyDeadband
Delivery amount at frame t: actualGenDA[sid, tt]

Equations (114) and (115) are constraint equations that constrain the relationship between the excess delivery amount indicator reOvGnDA[sid,t] and the excess delivery amount error genPlusDA[sid,t], which is 1 when the excess delivery amount error genPlusDA[sid,t] is positive and 0 otherwise. The variables related to equations (114) and (115) are as follows.

Excessive delivery error: genPlusDA[sid,t]
Predetermined value: bigM
Excessive delivery indicator: reOvGnDA[sid, t]

Equations (116) to (118) are constraint equations regarding the excess imbalance amount plusGenDA[sid,t], which takes the same value as the excess delivery amount error genPlusDA[sid,t] when the excess delivery amount indicator reOvGnDA[sid,t] is 1, and takes a value equal to or greater than zero so as to become 0 when the excess delivery amount indicator reOvGnDA[sid,t] is 0. The variables related to equations (116) to (118) are as follows.

Excessive delivery error: genPlusDA[sid,t]
Excess imbalance amount: plusGenDA[sid,t]
Predetermined value: bigM
Excessive delivery indicator: reOvGnDA[sid, t]

The equation (119) is a constraint equation for deriving the excess penalty penaltyPlsDA[sid,t] from the excess imbalance amount plusGenDA[sid,t]. The variables related to the equation (119) are as follows.

Excessive penalty: penaltyPlsDA[sid,t]
Penalty at frame t: ppnltOfDay[t]
Excess imbalance amount: plusGenDA[sid,t]

The equation (120) is a constraint equation for deriving the deficiency penalty penaltyMnsDA[sid,t] from the deficiency imbalance amount minusGenDA[sid,t]. The variables related to the equation (120) are as follows.


Deficient imbalance amount: minusGenDA[sid, t]
Penalty at frame t: ppnltOfDay[t]
Deficiency penalty: penaltyMnsDA[sid,t]

The formula (121) is a constraint formula for the delivery amount actTtlGenDA[sid, t] at each time t. The variables related to the formula (121) are as follows.

The value of the power generation day-ahead forecast series at time t: reGenDA[sid,t]
Charge amount of the storage battery at time t: actEsCfDA[sid, t]
Discharge amount of the storage battery at time t: actEsDfDA [sid, t]
Power output of private power generation equipment: actLclFDA [sid, t],
Power suppression amount of variable power supply: curtailDA[sid,t]
Battery charging efficiency: esEffc
Efficiency of the battery when discharging: esEffd

Equation (122) is a constraint on the change in the state of charge (SoC) of the storage battery. The variables related to equation (122) are as follows.

Storage battery rated capacity: ratedEsCapacity
Battery SoC at time t: esSocDA[sid,t]
Charge amount one time ago: actEsCfDA[sid, t-ctrlCycle]
Discharge amount one time ago: actEsDfDA [sid, t-ctrlCycle]
SoC at the previous time: esSocDA[sid, t-ctrlCycle]

Equation (123) is a constraint for ensuring that the charge amount of the storage battery is derived from the power (renewable energy power) of the variable power source 151. This equation is intended to keep the charge amount of the storage battery at time t equal to or less than the power generation power of the variable power source 151. The variables related to equation (123) are as follows.

Battery charging efficiency: esEffc
Charge amount of the storage battery at time t: actEsCfDA[sid, t]
The value of the power generation day-ahead forecast series at time tt: reGenDA[sid, tt]

Equation (124) is a constraint that prevents the charge amount actEsCfDA[sid,t] of the storage battery from exceeding the rated capacity ratedEsCapacity of the storage battery and the C value esCvC of the storage battery. The variables related to equation (124) are as follows.

Charge amount of the storage battery at time t: actEsCfDA[sid, t]
C value of storage battery: esCvC
Storage battery rated capacity: ratedEsCapacity
Charging execution indicator: esOcDA[sid, t]

Here, the optimization variable esOcDA[sid,t] is a charging indicator for time t, and is 1 if charging is in progress and 0 if not. When the charging indicator esOcDA[sid,t] is 1, the charge amount actEsCfDA[sid,t] of the storage battery at time t is upper bounded by a value determined by the rating and C value of the storage battery, but when it is 0, it is restricted to 0.

The formula (125) is a constraint equation similar to the formula (124) regarding the discharge amount actEsDfDA[sid,t] of the storage battery. The variables related to the formula (126) are as follows.

Discharge amount of the storage battery at time t: actEsDfDA [sid, t]
C value of battery discharge: esCvD
Storage battery rated capacity: ratedEsCapacity
Discharge execution indicator: esOdDA [sid, t]

Equation (126) and equation (127) are constraint equations expressing the charge amount actEsCfDA[sid,t] of the storage battery at time t. The variables related to equation (126) and equation (127) are as follows.

Charge amount of the storage battery at time t: actEsCfDA[sid, t]
Charging execution indicator: esOcDA[sid, t]

When the battery charge amount actEsCfDA[sid,t] is positive, the charging indicator esOcDA[sid,t] is constrained to 1 according to equation (126). On the other hand, when the battery charge amount actEsCfDA[sid,t] is 0, equation (126) holds true regardless of the charging indicator esOcDA[sid,t].

In contrast, in equation (127), when the charge amount actEsCfDA[sid,t] of the storage battery is sufficiently close to 0, the charge execution indicator esOcDA[sid,t] is constrained to 0. As a result, the charge amount actEsCfDA[sid,t] of the storage battery according to equation (124) is also constrained to 0.
When the absolute value of the charge amount actEsCfDA[sid,t] of the storage battery is made sufficiently large, the constraint on the charge execution indicator esOcDA[sid,t] does not function. Therefore, when the charge amount actEsCfDA[sid,t] of the storage battery is close to 0, the charge execution indicator esOcDA[sid,t] is constrained to 0, and otherwise is constrained to 1.

Equations (128) and (129) are constraint equations that express the relationship between the discharge amount actEsDfDA[sid,t] of the storage battery at time t and the discharge execution indicator esOdDA[sid,t]. The variables related to equation (128) are as follows.

Discharge amount of the storage battery at time t: actEsDfDA [sid, t]
Discharge execution indicator: esOdDA [sid, t]

When the battery discharge amount actEsDfDA[sid,t] is positive, the discharge execution indicator esOdDA[sid,t] is constrained to 1 according to equation (126). On the other hand, when the battery discharge amount actEsDfDA[sid,t] is 0, equation (126) holds true regardless of the discharge execution indicator esOdDA[sid,t].

In contrast, in equation (127), if the battery discharge amount actEsDfDA[sid,t] is sufficiently close to 0, the discharge execution indicator esOdDA[sid,t] is constrained to 0. As a result, according to equation (125), the battery discharge amount actEsDfDA[sid,t] is also constrained to 0.

On the other hand, when the absolute value of the discharge amount actEsDfDA[sid, t] of the storage battery is made sufficiently large, the restriction on the discharge execution indicator esOdDA[sid, t] does not function.
Therefore, when the discharge amount actEsDfDA[sid, t] of the storage battery is close to 0, the discharge execution indicator esOdDA[sid, t] is restricted to 0, and is restricted to 1 in other cases.
Equations (130) and (131) are constraints on opMaxEs, which is the upper operational limit SoC of the storage battery, and opMinEs, which is the lower operational limit SoC of the storage battery.

Equation (130) is a constraint that restricts the operation upper limit from being exceeded even if the battery is charged from the SoC of the battery at time t, esSocDA[sid,t], at the charge amount actEsCfDA[sid,t] of the battery over the control period ctrlCycle. The variables related to equation (130) are as follows.

Charge amount of the storage battery at time t: actEsCfDA[sid, t]
Battery SoC at time t: esSocDA[sid,t]
Storage battery rated capacity: ratedEsCapacity
Battery operation limit SoC: opMaxEs

Equation (131) is a constraint that restricts the SoC of the storage battery at time t from falling below the lower operational limit even if the storage battery is discharged for a period of ctrlCycle at the discharge amount actEsDfDA[sid,t] of the storage battery at time t. The variables related to equation (131) are as follows.

Discharge amount of the storage battery at time t: actEsDfDA [sid, t]
Storage battery rated capacity: ratedEsCapacity
Battery SoC at time t: esSocDA[sid,t]
Battery operation lower limit SoC: opMinEs

Equation (132) is a constraint for preventing simultaneous charging and discharging of the storage battery. The variables related to equation (132) are as follows:

Charging execution indicator: esOcDA[sid, t]
Discharge execution indicator: esOdDA [sid, t]

Equations (133) and (134) are constraints that restrict the power generation output actLclFDA[sid,t] of the private power generation facility to the range between the rated output ratedLocalFlex and the minimum operating output minLocalFlex.

Equation (133) indicates that the product of the minimum operating output minLocalFlex and the operation indicator lfODA[sid,t] of the private power generation facility is the minimum value of the power generation output actLclFDA[sid,t] of the private power generation facility. The variables related to equation (133) are as follows.

Minimum operating output: minLocalFlex
Operation indicator of private power generation equipment: lfODA [sid, t]
Power output of private power generation equipment: actLclFDA [sid, t]

Equation (134) indicates that the product of the rated output ratedLocalFlex and the operation indicator lfODA[sid,t] of the private power generation facility is the maximum value of the power generation output actLclFDA[sid,t] of the private power generation facility. The variables related to equation (134) are as follows.

Power output of private power generation equipment: actLclFDA [sid, t]
Rated output: ratedLocalFlex
Operation indicator of private power generation equipment: lfODA [sid, t]

The operation indicator lfODA[sid,t] of the private power generation facility is a variable that is set to 1 when the private power generation facility is operated and set to 0 when the private power generation facility is stopped. When the operation indicator lfODA[sid,t] of the private power generation facility is set to 1, the power generation output actLclFDA[sid,t] of the private power generation facility is constrained between the minimum output and the rated value. When the operation indicator lfODA[sid,t] of the private power generation facility is set to 0, the power generation output actLclFDA[sid,t] of the private power generation facility is constrained to 0.

Equation (135) is a constraint for avoiding charging the storage battery with energy derived from a private power generation facility. The variables related to equation (135) are as follows.

Charging execution indicator: esOcDA[sid, t]
Operation indicator of private power generation equipment: lfODA [sid, t]

By generating such optimization variables and constraint equations and providing them to the optimization engine 170, it is possible to generate the bid amount reBidsDA[t] at frame t that maximizes the profit, which is the objective function, and the operation command series for the variable power source 151 and the controllable power source 152, such as actEsCfDA[sid, tt], actEsDfDA[sid, tt], curtailDA[sid, tt], and actLclFDA[sid, tt] that realizes this.

The day-ahead bidding planning means 103 sets the selling bid price to the bid amount reBidsDA[t] for each frame t obtained in this way and submits a bid to the day-ahead market. This allows the power selling control device 100 to submit a bid so that the power generated by the power generation site 150 can be sold appropriately, and control the power generation sites 150 and 150a according to the contracted amount of the bid.

Next, we will explain the processing of the intraday bidding planning means 107, which is implemented using the agreement result 114 based on the bids obtained by a specified timing, the intraday power generation forecast series 106 which is the output of the intraday power generation time advance forecasting means 105, and the output 116 and state 117 of the variable power source 151 and the output 118 and state 119 of the controllable power source 152.

Equation (201) represents an example of an objective function for optimization in the daily plan.

Revenue from electricity market sales: earningsHA[sid,t]
Premium: premiumHA[sid,t]
Planning period: planHorizon
Operating cost of self-power generation equipment: lflexXpnsHA[sid,t]
Penalty: penaltyHA[sid,t]
Cost of electricity procured in the market today: pcsGenXpnsHA[sid,t]

In the daily plan, the objective function is to maximize the average value of the revenue for each daily power generation forecast series 106 identified by sid during the planning period minus the total value during the planning period, that is, the average value for the entire daily power generation forecast series 106.
The optimization variables in the daily plan are the following five sets.

Additional electricity selling bid amount for each frame t on the day: reBidsHA[t]
Adjustment amount procured in the market on the day: pcsGenHA [t]
Power suppression amount of variable power supply at time tt: curtailHA[sid, tt]
Battery charge amount: actEsCfHA[sid,t]
Battery discharge amount: actEsDfHA [sid, tt]
Power generation output of private power generation equipment: actLclFHA [sid, t]

Equation (202) is a constraint on the adjustment amount pcsGenHA[t] procured in the market on the day. The variables related to equation (202) are as follows.

Cost of electricity procured in the market today: pcsGenXpnsHA[sid,t]
Today's market gain: MktGain
Predicted market price in the current day: mktPrcOfHA[t]
Adjustment amount procured in the market on the day: pcsGenHA [t]

Since the intraday market is a continuous session, it is considered that price prediction itself is difficult. Here, the unit price is calculated by multiplying the predicted market price mktPrcOfHA[t] in the intraday market based on the predicted value of the previous day's market price and the flash value by the gain MktGain in the intraday market, and the cost of the adjustment amount pcsGenHA[t] procured in the intraday market is calculated, and the cost of electricity procured in the intraday market pcsGenXpnsHA[sid,t] is constrained.

Similarly to equation (102), equation (203) is a constraint equation for revenues earnedsHA[sid,t] from electricity market sales. The variables related to equation (203) are as follows.

Revenue from electricity market sales: earningsHA[sid,t]
Predicted market price in the current day: mktPrcOfHA[t]
Delivery amount at frame t: actualGenHA[sid, tt]

Equation (204) is a constraint equation for the penalty HA[sid, t] similar to equation (103). The variables related to equation (204) are as follows.

Penalty: penaltyHA[sid,t]
Excess penalty at frame t: penaltyPlsHA[sid, tt]
Deficiency penalty at piece t: penaltyMnsHA[sid, tt]

Equation (205) is a constraint equation for the premium premiumHA[sid,t] similar to equation (104). The variables in equation (205) are as follows.

Premium: premiumHA[sid,t]
Predicted or expected value of premium unit price: prmOfDay[t]
The value of the power generation day-ahead forecast series at time tt: reGenHA[sid, tt]
Efficiency of the battery when discharging: esEffd
Discharge amount of the storage battery at time tt: actEsFlexD[sid, tt]
Battery charging efficiency: esEffc
Charge amount of the storage battery at time tt: actEsFlexC[sid, tt]
Power suppression amount of variable power supply: curtailHA[sid,t]

The formula (206) is a constraint formula for the operating cost lflexXpnsHA[sid,t] of the private power generation facility, similar to the formula (105). The variables related to the formula (206) are as follows.

Operating cost of the self-power generation equipment at t: lflexXpnsHA[sid,t]
Operating cost of private power generation equipment: localFexPrice
Power generation output of the private power generation facility within time tt: actLclFHA [sid, tt]

Equation (207) is a constraint on the delivery amount actualGenHA[sid,t] obtained by dividing the frame into one or more parts, similar to equation (106). The variables related to equation (207) are as follows.

Amount of delivery at frame t: actualGenHA[sid,t]
Amount transferred at time tt: actTtlGenHA[sid, tt]

Equations (208) to (213) define the error between the amount of the contract amount for that day's bids, taking into account the amount that has been contracted, minus the delivery amount deadband width pnltyDeadband, and the delivery amount, which corresponds to equations (107) to (112) for the previous day's plan, and defines the delivery amount shortage error genMnusHA[sid,t], which is positive when there is a shortage.

Equation (208) defines the delivery amount shortage error genMnusHA[sid, t] that is positive when the delivery amount is insufficient. The variables related to equation (208) are as follows.

Delivery amount insufficiency error: genMnusHA[sid, t]
Bid volume at frame t on the day-ahead market: reBidsDA[t]
Bid volume in the market for today's time slot t: reBidsHA[t]
Delivery amount deadband width: pnlteyDeadband
Delivery amount at frame t: actualGenHA[sid, tt]

The contract amount reBidsDA[t] of the day-ahead bid is information included in the contract result 114 notified from the day-ahead market system 120, and is treated as a constant in the optimization calculation performed by the intraday bid planning means 107.

Furthermore, the portion reBidsHA[t] relating to intraday bids, for the time when the gate was closed, is an aggregate of information contained in the contract results 115 notified by the intraday market system 130. For time t before the gate was closed, if there are contract results 115 for time t, the portion is formulated as the sum of the values thereof and an optimization variable for planning regarding additional bid volume.

Equations (209) and (210) are constraints corresponding to equations (108) and (109) in the day-ahead plan. Equations (209) and (210) are constraints that constrain the relationship between the delivery amount shortage indicator reUdGnHA[sid,t] and the delivery amount shortage error genMnusHA[sid,t], which is 1 when the delivery amount shortage error genMnusHA[sid,t] is positive and 0 otherwise. The variables related to equations (209) and (210) are as follows:

Delivery amount insufficiency error: genMnusHA[sid, t]
Predetermined value: bigM
Delivery amount shortage indicator: reUdGnHA[sid, t]

When the delivery amount insufficiency error genMnusHA[sid,t] is positive, formula (209) holds regardless of the delivery amount insufficiency indicator reUdGnHA[sid,t] and does not impose any constraint, but formula (210) constrains the delivery amount insufficiency indicator reUdGnHA[sid,t] to 1. On the other hand, when the delivery amount insufficiency error genMnusHA[sid,t] is negative, formula (210) holds regardless of the delivery amount insufficiency indicator reUdGnHA[sid,t] and does not impose any constraint, but formula (209) constrains the delivery amount insufficiency indicator reUdGnHA[sid,t] to 0.

Equations (211) to (213) are constraint equations related to the deficiency imbalance amount minusGenHA[sid,t], which takes the same value as the delivery amount deficiency error genMnusHA[sid,t] when the delivery amount deficiency indicator reUdGnHA[sid,t] is 1, and takes a value equal to or greater than zero such that it becomes 0 when the delivery amount deficiency indicator reUdGnHA[sid,t] is 0. The variables related to equations (211) to (213) are as follows.

Delivery amount insufficiency error: genMnusHA[sid, t]
Deficient imbalance amount: minusGenHA[sid, t]
Delivery amount shortage indicator: reUdGnHA[sid, t]

Equations (211) and (212) correspond to equations (101) and (111) of the day-ahead plan, and the shortage imbalance amount minusGenHA[sid,t] is constrained to the shortage error genMnusHA[sid,t] of the delivery amount when the delivery amount shortage indicator reUdGnHA[sid,t] is 1. When the delivery amount shortage indicator reUdGnHA[sid,t] is 0, the shortage imbalance amount minusGenHA[sid,t] is constrained to 0 by equation (213) corresponding to equation (112) of the day-ahead plan.

Equations (214) to (219) correspond to equations (113) to (118) in the day-ahead plan. Equations (214) to (219) are constraint equations that constrain the relationship between the delivery amount excess error genPlusHA[sid,t], the delivery amount excess indicator reOvGnHA[sid,t], and the excess imbalance amount plusGenHA[sid,t] in the same manner as in the case of equations (208) to (213). The variables related to equations (214) to (219) are as follows.

Excessive delivery error: genPlusHA[sid,t]
Delivery amount at frame t: actualGenHA[sid, tt]
Bid volume in the market for today at time t: reBidsDA[t]
Bid volume at frame t on the previous day's market: reBidsHA[t]
Delivery amount deadband width: pnlteyDeadband
Excessive delivery amount indicator: reOvGnHA[sid, t]
Excess imbalance amount: plusGenHA[sid,t]

The equation (220) is a constraint equation for deriving the excess penalty penaltyPlsHA[sid,t] from the excess imbalance amount plusGenHA[sid,t]. The variables related to the equation (220) are as follows.

Excessive penalty: penaltyPlsHA[sid,t]
Excess imbalance amount: plusGenHA[sid,t]

The equation (221) is a constraint equation for deriving the deficiency penalty penaltyMnsHA[sid,t] from the deficiency imbalance amount minusGenHA[sid,t]. The variables related to the equation (221) are as follows.

Deficient imbalance amount: minusGenHA[sid, t]
Deficiency penalty: penaltyMnsHA[sid,t]

Equation (222) is a constraint equation for the delivery amount actTtlGenHA[sid,t] at time t, which is the same constraint as equation (121) in the day-ahead plan. The variables related to equation (222) are as follows.

The predicted value of the power generation amount of the variable power source at time t on the day: reGenHA[sid,t]
Power suppression amount of variable power supply at control time t: curtailHA[sid, t]
Adjustment amount procured in the market on the day: pcsGenHA [t]
Battery charge amount: actEsCfHA[sid,t]
Discharge amount of storage battery: actEsDfHA [sid, t]
Power generation output of private power generation equipment: actLclFHA [sid, t]

The adjustment capacity to be procured from the market is defined for the time during the planning period planHorizon when the market gate is not closed on the day, but for the time period before that, the contract result 115 that is contracted and notified by the intraday market system 130 is used as a constant.

The following equations (223) to (233) are constraint equations related to the storage battery in the hour-ahead plan. The equations (223) to (233) correspond to the equations (122) to (132) in the day-ahead plan.

Equation (223) corresponds to equation (122) in the day-ahead plan, and is a constraint on the change in the state of charge (SoC) of the storage battery. The variables related to equation (223) are as follows.

Storage battery rated capacity: ratedEsCapacity
The SoC of the storage battery at time t: esSocHA[sid,t]
Charge amount of the storage battery one time ago: actEsCfHA [sid, t-ctrlCycle]
Discharge amount of the storage battery one time ago: actEsDfHA [sid, t-ctrlCycle]
SoC at the previous time: esSocHA[sid, t-ctrlCycle]

Equation (224) corresponds to equation (123) in the day-ahead plan, and is a constraint to ensure that the discharge power from the storage battery is derived from the power of the variable power source (renewable energy power). It is to suppress the charge amount at time t to be equal to or less than the power generation power of the variable power source. The variables related to equation (224) are as follows.

Battery charging efficiency: esEffc
Charge amount of the storage battery at time t: actEsCfHA[sid, t]
Power suppression amount of variable power supply: curtailHA[sid,t]
The predicted value of the power generation amount of the variable power source at time t on the day: reGenHA[sid,t]

The formula (225) corresponds to the formula (124) in the day-ahead plan, and is a constraint that the charge amount actEsCfHA[sid,t] of the storage battery does not exceed the rated capacity ratedEsCapacity of the storage battery and the C value esCvC of the storage battery. The variables related to the formula (225) are as follows.

Charge amount of the storage battery at time t: actEsCfHA[sid, t]
C value of storage battery: esCvC
Storage battery rated capacity: ratedEsCapacity
Charging execution indicator: esOcHA[sid, t]

Equation (226) corresponds to equation (125) in the day-ahead plan, and is a constraint equation similar to equation (225) regarding the optimization variable actEsDfHA[sid, t] representing the discharge power. The variables related to equation (226) are as follows.

Discharge amount of the storage battery at time t: actEsDfHA[sid, t]
C value of battery discharge: esCvD
Storage battery rated capacity: ratedEsCapacity
Discharge execution indicator: esOdHA[sid,t]

Equations (227) and (228) correspond to equations (126) and (127) in the day-ahead schedule, and are constraint equations that represent the charge amount actEsCfHA[sid,t] of the storage battery at time t. The variables related to equations (227) and (228) are as follows.

Charge amount of the storage battery at time t: actEsCfHA[sid, t]
Charging execution indicator: esOcHA[sid, t]

Equations (229) and (230) correspond to equations (128) and (129) in the day-ahead plan, and are constraint equations that represent the relationship between the optimization variable actEsDfHA[sid,t] of the discharge power of the storage battery at time t and the discharge execution indicator esOdHA[sid,t]. The variables related to equations (229) and (230) are as follows.

Discharge amount of the storage battery at time t: actEsDfHA[sid, t]
Discharge execution indicator: esOdHA[sid,t]

Equation (231) corresponds to equation (130) in the day-ahead plan, and is a constraint that restricts the SoC of the storage battery at time t, esSocHA[sid,t], from exceeding the upper operational limit SoC of the storage battery even if the storage battery is charged at the charge amount actEsCfHA[sid,t] over the control period ctrlCycle. The variables related to equation (231) are as follows.

Charge amount of the storage battery at time t: actEsCfHA[sid, t]
The SoC of the storage battery at time t: esSocHA[sid,t]
Storage battery rated capacity: ratedEsCapacity
Battery operation limit SoC: opMaxEs

Equation (232) corresponds to equation (131) in the day-ahead plan, and is a constraint that restricts the current SoC from falling below the lower operational limit SoC of the battery even if the battery is discharged for a period of ctrlCycle at the discharge amount actEsDfHA[sid,t]. The variables related to equation (232) are as follows.

Discharge amount of the storage battery at time t: actEsDfHA[sid, t]
Storage battery rated capacity: ratedEsCapacity
The SoC of the storage battery at time t: esSocHA[sid,t]
Battery operation lower limit SoC: opMinEs

The formula (233) corresponds to the formula (132) in the day-ahead plan, and is a constraint to prevent simultaneous charging and discharging. The variables related to the formula (233) are as follows.

Charging execution indicator: esOcHA[sid, t]
Discharge execution indicator: esOdHA[sid,t]

Equations (234) and (235) are constraints on the operation of the private power generation equipment in the hour-ahead planning. They correspond to equations (133) and (134) in the day-ahead planning.

Equation (234) indicates that the product of the minimum operating output minLocalFlex and the operation indicator lfOHA[sid,t] of the private power generation facility is the minimum value of the power generation output actLclFHA[sid,t] of the private power generation facility. The variables related to equation (234) are as follows.

Minimum operating output: minLocalFlex
Operation indicator of private power generation equipment: lfOHA [sid, t]
Power generation output of private power generation equipment: actLclFHA [sid, t]

Equation (235) indicates that the product of the rated output ratedLocalFlex and the operation indicator lfOHA[sid,t] of the private power generation facility is the maximum value of the power generation output actLclFHA[sid,t] of the private power generation facility. The variables related to equation (235) are as follows.

Power generation output of private power generation equipment: actLclFHA [sid, t]
Rated output: ratedLocalFlex
Operation indicator of private power generation equipment: lfOHA [sid, t]

Equation (236) is a constraint for avoiding storing energy derived from a private power generation facility in the power to be charged to the storage battery in the hour-ahead plan, and corresponds to equation (135) in the day-ahead plan. The variables related to equation (236) are as follows.

Charging execution indicator: esOcHA[sid, t]
Operation indicator of private power generation equipment: lfOHA [sid, t]

Equations (237) to (242) are constraint equations for the indicator esOcHc[sid,t] indicating a change in the charging state of the storage battery, the indicator esOdHc[sid,t] indicating a change in the discharging state, and the indicator lfxOHc[sid,t] indicating a change in the operating state of the private power generation equipment.
Equations (237) to (242) are constraints on the difference between the charging execution indicator esOcHA[sid,t] representing the charging operation of the storage battery, the discharging execution indicator esOdHA[sid,t] representing the discharging operation of the storage battery, and the operation indicator lfOHA[sid,t] of the private power generation equipment, respectively.

For example, consider the indicator esOcHc[sid, t] and the charging execution indicator esOcHA[sid, t].
From equation (237), when charging operation starts, esOcHA[sid,t-ctrlCycle]=0, and esOcHA[sid,t]=1. At this time, esOcHA[sid,t]≧1, and the indicator esOcHc[sid,t] on the left side is restricted to 1. At other times, the indicator esOcHc[sid,t] is not restricted to either 0 or 1.

On the other hand, when esOcHA[sid,t] and esOcHA[sid,t-ctrlCycle] are equal, the left side of equation (238) is constrained to be equal to or greater than 1. In order for this constraint to hold, esOcHc[sid,t] needs to be 0, so the indicator esOcHc[sid,t] can be constrained to be 1 only at the start of charging by the two constraint equations (237) and (238).

Similarly, for equations (239) to (242), the indicator esOdHc[sid, t] is constrained to be 1 only when discharge starts, and lfOHc[sid, t] is constrained to be 1 only when the private power generation facility starts operating.

Equations (243) to (245) are constraints that these indicators do not exceed a certain number onoffmax within a certain period.

By generating such optimization variables, constraint equations, and objective functions and providing them to the optimization engine 170 in FIG. 1, it is possible to generate the additional power selling bid amount reBidsHA[t] for each frame t on the day that maximizes the profit, which is the objective function, the adjustment amount pcsGenHA[t] to be procured in the market on the day, and operation command sequences for the variable power sources 151 and controllable power sources 152, such as actEsCfHA[sid, tt], actEsDfHA[sid, tt], curtailHA[sid, tt], and actLclFHA[sid, tt] that realize this.

The intraday bidding planning means 107 sets a selling bid price for the additional electricity selling bid amount reBidsHA[t] for each frame t on the day obtained in this way, and a buying bid price for the adjustment amount pcsGenHA[t] to be procured in the intraday market, and submits the bid to the intraday market system 130. This allows the electricity selling control device 100 to submit bids so that the electricity generated by the power generation site 150 can be sold appropriately, and control the power generation sites 150, 150a according to the contracted amount of the bid.

Next, the process of the operation planning means 111 will be described, which is performed using the contract result 115 for the current day's buying and selling bid amount 108 obtained by a predetermined timing, the contract result 114 for the previous day's bid amount 104, and the power generation forecast time series 110 of the power generation high frequency short-term forecasting means 109. Note that the forecast frequency of the power generation high frequency short-term forecasting means 109 is preferably equal to or several times the control cycle ctrlCycle. The forecast period needs to be the period up to the end of the frame or longer from the start of the frame (for example, the time when the control command for 13:00 is decided), but it is not necessary to forecast up to one day or several hours ahead, as in the day-ahead bid planning means 103 and the current day bid planning means 107.

Equation (301) is an objective function of the optimization performed by the operation planning means 111. This is a minimization problem of the optimization variable costRt[t] defined by equation (302), and the following optimization variables are determined.
Battery charge amount: actEsCfRT [tt]
Discharge amount of storage battery: actEsDfRT [tt]
Amount of power suppression of variable power supply: curtailRT [tt]
Power output of private power generation equipment: actLclFRT [tt]

The day-ahead bid planning means 103 and the day-ahead bid planning means 107 perform stochastic optimization using a plurality of predicted power generation amount series, but the operation planning means 111 of this embodiment uses one series for one planning process. That is, the power generation forecast time series 110 based on the latest forecast for each plan is used in a plan in which time t is the first time of the planning variables and a plan in which time t + ctrlCycle is the first time of the planning target.

In the planning process, the period of the integration process in equation (301) is t ∈ Ctrl Horizon, and at least one frame is the subject of the planning process.

The right-hand side of formula (302) is largely composed of two terms. The first term is a term weighted by weightOfActualGen, a setting parameter that takes a value between 0 and 1, and is a term for the imbalance cost for the excess or deficiency of the delivery amount.

Predicted or estimated value of excess imbalance penalty unit price at frame t0: ppnltOfDay[t0]
Excessive imbalance amount: genErrPRT [t]
Predicted or expected value of the shortage imbalance penalty at frame t0: mpnltOfDay[t0]
Deficiency imbalance amount: genErrMRT [t]

The second term is a term weighted by (1-weightOfActualGen) and is a term representing the cost of the renewable energy suppression amount, the charge amount of the storage battery, the discharge amount of the storage battery, and the power generation output of the private power generation facility.

Cost of suppressing power from variable power sources: curtailCost
Amount of power suppression of variable power supply: curtailRT [tt]
Battery charging cost: chargeCost
Battery charge amount: actEsCfRT [tt]
Battery discharge cost: dischargeC
Discharge amount of storage battery: actEsDfRT [tt]
Operating cost of private power generation equipment: localFexPrice
Power output of private power generation equipment: actLclFRT [tt]

For example, the power suppression amount curtailRT[tt] of a variable power source is an optimization variable defined for the control timing tt, and the cost is calculated by adding the value of the variable defined between the start time t0 minutes and t0+30 minutes of the target frame multiplied by ctrlCycle/60 and converted into power energy, and then multiplying the sum by the unit price.

The possible values of the target time of the planning variable are one of the discrete values shown in Figure 3. In Figure 3, the target time is shown as [t0, t0 + ctrlCycle, t0 + 2 * ctrlCycle, ..., t, ..., t0 + div (30, ctrlCycle, 0) * ctrlCycle].

Of these, the part of the planning variable before the first time t is in the past, so it is not treated as a variable, and actual values are used. The part after time t is generated as an optimization variable and used to form a constraint equation.

The same applies to the charge amount actEsCfRT[tt] of the storage battery, the discharge amount actEsDfRT[tt] of the storage battery, and the power output actLclFRT[t] of the private power generation facility. On the other hand, for the past part, the actual value is used to function as a constraint value. Here, div(a, b, c) means the quotient with C decimal places when a is divided by b.

Formula (303) is a constraint equation that constrains the difference between the transfer amount actTtlGenRT[t] at the target frame (time t) and the sum of the contract amount reBidsDA[t] in the previous day's bid and the contract amount reBidsHA[t] in the current day's bid to the transfer amount error genErrorRT[t]. The variables related to formula (303) are as follows.

Delivery amount error: genErrorRT[t]
Amount transferred at time tt: actTtlGenRT[tt]
Contract amount on the previous day's bid: reBidsDA[t]
Contract amount on the day of bidding: reBidsHA[t]

Equations (304) and (305) are constraint equations that constrain the excess imbalance amount genErrPRT[t] to the delivery amount error genErrorRT[t] when the delivery amount excess indicator flgGePRT[t] is 1, and are essentially unconstrained when the delivery amount excess indicator flgGePRT[t] is 0. The variables related to equations (304) and (305) are as follows:

Excessive imbalance amount: genErrPRT[sid, t]
Excessive delivery indicator: flgGePRT[sid, t]
Delivery amount error: genErrorRT[sid, t]

When the delivery amount excess indicator flgGePRT[t] is 1, the formula (306) does not function as a constraint on the excess imbalance amount genErrPRT[t]. However, when the delivery amount excess indicator flgGePRT[t] is 0, the excess imbalance amount genErrPRT[t] is constrained to 0. The variables related to the formula (306) are as follows.

Excessive delivery indicator: flgGePRT[t]
Excessive imbalance amount: genErrPRT [t]

Equations (307) and (308) are constraint equations regarding the relationship between the delivery amount error genErrorRT[t] and the delivery amount excess indicator flgGePRT[t]. The variables related to equations (307) and (308) are as follows.

Delivery amount error: genErrorRT[t]
Excessive delivery indicator: flgGePRT[t]

When the value of the delivery amount error genErrorRT[t] is positive, equation (308) holds regardless of the delivery amount excess indicator flgGePRT[t]. However, equation (307) restricts the delivery amount excess indicator flgGePRT[t] to 1.
On the other hand, when the delivery amount error genErrorRT[t] is negative, equation (307) holds regardless of the value of the delivery amount excess indicator flgGePRT[t]. However, equation (308) restricts the delivery amount excess indicator flgGePRT[t] to 0.

Equations (309) and (310) are constraint equations that constrain the imbalance shortage amount genErrMRT[t] to the inverted sign of the delivery amount error genErrorRT[t] when the delivery amount shortage indicator flgGeMRT[t] is 1, and are substantially unconstrained when the delivery amount shortage indicator flgGeMRT[t] is 0. The variables related to equations (309) and (310) are as follows.

Deficiency imbalance amount: genErrMRT [t]
Delivery amount error: genErrorRT[t]
Delivery amount shortage indicator: flgGeMRT [t]

When the delivery amount shortage indicator flgGeMRT[t] is 1, the formula (311) does not function as a constraint on the imbalance shortage amount genErrMRT[t]. However, when the delivery amount shortage indicator flgGeMRT[t] is 0, the imbalance shortage amount genErrMRT[t] is constrained to 0. The variables related to the formula (311) are as follows.


Delivery amount shortage indicator: flgGeMRT [t]
Deficiency imbalance amount: genErrMRT [t]

Equation (312) and equation (313) are constraint equations regarding the relationship between the delivery amount error genErrorRT[t] and the delivery amount shortage indicator flgGeMRT[t]. The variables related to equation (312) and equation (313) are as follows.

Delivery amount error: genErrorRT[t]
Delivery amount shortage indicator: flgGeMRT [t]

When the delivery amount error genErrorRT[t] is negative, equation (313) holds regardless of the value of the delivery amount shortage indicator flgGeMRT[t]. However, equation (312) restricts the delivery amount shortage indicator flgGeMRT[t] to 1.
On the other hand, when the delivery amount error genErrorRT[t] is positive, equation (312) holds regardless of the value of the delivery amount deficiency indicator flgGeMRT[t], but equation (313) restricts the value of the delivery amount deficiency indicator flgGeMRT[t] to 0.

From equation (314), the sum of the delivery amount excess indicator flgGePRT[t] and the delivery amount deficiency indicator flgGeMRT[t] is restricted to be equal to or less than 1. The variables related to equation (314) are as follows.

Excessive delivery indicator: flgGePRT[t]
Delivery amount shortage indicator: flgGeMRT [t]

In the first term on the right side of formula (302), these values are used to calculate the penalty for the excess or shortage. Note that ppnltOfDay[t0] in formula (302) is a predicted or assumed value of the excess imbalance penalty unit price for frame t0, and mpnltOfDay[t0] is a predicted or assumed value of the shortage imbalance penalty for frame t0.

Equation (315) is a constraint equation for the optimization variables expressing the delivery amount actTtlGenRT[t] at time t used in equation (303).

Amount transferred at time t: actTtlGenRT[t]
Power generation forecast time series value: reGenRT[t]
Power suppression amount of variable power supply: curtailRT [t]
Amount of purchased electricity agreed on the day's market: pchsGenHA [t]
Battery charging efficiency: esEffc
Charge amount of storage battery: actEsCfRT [t]
Battery charging efficiency: esEffc
Discharge amount of storage battery: actEsDfRT [t]
Power generation output of private power generation equipment: actLclFRT [t]

The power generation forecast time series 110 is an output of the power generation high frequency short-term forecasting means 109 .

By generating such optimization variables, constraint equations, and objective functions and providing them to the optimization engine 170 in FIG. 1, it is possible to generate control command sequences for the variable power source 151 and the controllable power source 152, such as actEsCfRT[tt], actEsDfRT[tt], curtailRT[tt], and actLclFRT[tt], which minimize the cost, which is the objective function.

Of the calculation results in the operation planning means 111, the optimization result for the first time t of the planning target is set as a control command series to the variable power source 151 and the controllable power source 152. Since the private power generation equipment and the storage battery shown in this embodiment are controllable power sources 152, actEsCfRT[t], actEsDfRT[t], and actLclFRT[t] are transmitted by the command means 166 as command values 113 to the controllable power sources 152.

Since curtailRT[t] is for the variable power source 151, it is transmitted by the command means 165 as the command value 112 for the variable power source 151. This allows the power selling control device 100 to make a bid so that the power generated by the power generation site 150 can be sold appropriately, and to control the power generation sites 150, 150a according to the contract amount of the bid.

Second Embodiment
In the first embodiment of the present invention, it is assumed that the high-frequency short-term power generation prediction means 109 generates one power generation prediction series frequently, but it may generate multiple prediction series like the power generation time-before prediction means 105 and the power generation day-before prediction means 101. In that case, the objective function and constraint equations of equations (401) to (415) may be used instead of the objective function and constraint equations of equations (301) to (315).

Equation (401) is an objective function of the optimization performed by the operation planning means 111. This is a minimization problem of the optimization variable costRt[t] defined by equation (402), and the following optimization variables are determined.

Battery charge amount: actEsCfRT[sid, tt]
Discharge amount of storage battery: actEsDfRT [sid, tt]
Power suppression amount of variable power supply: curtailRT [sid, tt]
Power output of private power generation equipment: actLclFRT [sid, tt]

The right-hand side of formula (402) is largely composed of two terms. The first term is a term weighted by weightOfActualGen, a setting parameter that takes a value between 0 and 1, and is a term for the imbalance cost for the excess or deficiency of the delivery amount.

Predicted or estimated value of excess imbalance penalty unit price at frame t0: ppnltOfDay[t0]
Excessive imbalance amount: genErrPRT [sid, tt]
Predicted or expected value of the shortage imbalance penalty at frame t0: mpnltOfDay[t0]
Deficiency imbalance amount: genErrMRT [sid, tt]

The second term is a term weighted by (1-weightOfActualGen) and is a term representing the cost of the power generation suppression amount of the variable power source, the charge amount of the storage battery, the discharge amount of the storage battery, and the power generation amount of the private power generation facility.

Cost of suppressing power from variable power sources: curtailCost
Amount of power generation suppression of variable power source: curtailRT [sid, tt]
Battery charging cost: chargeCost
Battery charge amount: actEsCfRT[sid, tt]
Battery discharge cost: dischargeC
Discharge amount of storage battery: actEsDfRT [sid, tt]
Operating cost of private power generation equipment: localFexPrice
Power output of private power generation equipment: actLclFRT [sid, tt]

Formula (403) is a constraint equation that constrains the difference between the transfer amount actTtlGenRT[sid, tt] at time t, which is the target frame, and the sum of the contract amount reBidsDA[t] in the previous day's bid and the contract amount reBidsHA[t] in the current day's bid, to the transfer amount error genErrorRT[sid, t]. The variables related to formula (403) are as follows.

Delivery amount error: genErrorRT[sid, t]
Amount transferred at time t: actTtlGenRT[sid, tt]
Contract amount on the previous day's bid: reBidsDA[t]
Contract amount on the day of bidding: reBidsHA[t]

Equations (404) and (405) are constraint equations that constrain the excess imbalance amount genErrPRT[sid,t] to the delivery amount error genErrorRT[sid,t] when the delivery amount excess indicator flgGePRT[sid,t] is 1, and that substantially unconstrain the excess imbalance amount genErrPRT[sid,t] when the delivery amount excess indicator flgGePRT[sid,t] is 0. The variables related to equations (404) and (405) are as follows:

Excessive imbalance amount: genErrPRT[sid, t]
Excessive delivery indicator: flgGePRT[sid, t]
Delivery amount error: genErrorRT[sid, t]

When the delivery amount excess indicator flgGePRT[sid,t] is 1, the formula (406) does not function as a constraint on the excess imbalance amount genErrPRT[sid,t]. However, when the delivery amount excess indicator flgGePRT[sid,t] is 0, the excess imbalance amount genErrPRT[sid,t] is constrained to 0. The variables related to the formula (406) are as follows.

Excessive delivery indicator: flgGePRT[sid, t]
Excessive imbalance amount: genErrPRT[sid, t]

Equations (407) and (408) are constraint equations regarding the relationship between the delivery amount error genErrorRT[sid,t] and the delivery amount excess indicator flgGePRT[sid,t]. The variables related to equations (407) and (408) are as follows.


Delivery amount error: genErrorRT[sid, t]
Excessive delivery indicator: flgGePRT[sid, t]

Equations (409) and (410) are constraint equations that constrain the imbalance deficiency amount genErrMRT[sid,t] to the inverted sign of the delivery amount error genErrorRT[sid,t] when the delivery amount deficiency indicator flgGeMRT[sid,t] is 1, and are substantially unconstrained when the delivery amount deficiency indicator flgGeMRT[sid,t] is 0. The variables related to equations (409) and (410) are as follows.

Deficiency imbalance amount: genErrMRT [sid, t]
Delivery amount error: genErrorRT[sid, t]
Delivery amount shortage indicator: flgGeMRT[sid, t]

When the delivery amount shortage indicator flgGeMRT[sid,t] is 1, the formula (411) does not function as a constraint on the shortage imbalance amount genErrMRT[sid,t]. However, when the delivery amount shortage indicator flgGeMRT[sid,t] is 0, the shortage imbalance amount genErrMRT[sid,t] is constrained to 0. The variables related to the formula (411) are as follows.

Delivery amount shortage indicator: flgGeMRT[sid, t]
Deficiency imbalance amount: genErrMRT [sid, t]

Equations (412) and (413) are constraint equations regarding the relationship between the delivery amount error genErrorRT[sid,t] and the delivery amount shortage indicator flgGeMRT[sid,t]. The variables related to equations (412) and (413) are as follows.

Delivery amount error: genErrorRT[sid, t]
Delivery amount shortage indicator: flgGeMRT[sid, t]

When the delivery amount error genErrorRT[sid,t] is negative, equation (413) holds regardless of the value of the delivery amount deficiency indicator flgGeMRT[sid,t]. However, equation (412) restricts the delivery amount deficiency indicator flgGeMRT[sid,t] to 1.
On the other hand, when the delivery amount error genErrorRT[sid,t] is positive, equation (412) holds regardless of the value of the delivery amount deficiency indicator flgGeMRT[sid,t], but equation (413) restricts the value of the delivery amount deficiency indicator flgGeMRT[sid,t] to 0.

From equation (414), the sum of the delivery amount excess indicator flgGePRT[sid,t] and the delivery amount deficiency indicator flgGeMRT[sid,t] is restricted to be equal to or less than 1. The variables related to equation (414) are as follows.

Excessive delivery indicator: flgGePRT[sid, t]
Delivery amount shortage indicator: flgGeMRT[sid, t]

The formula (415) is a constraint formula for the delivery amount actTtlGenRT[sid, t] at time t used in the formula (403). The variables related to the formula (415) are as follows.

Delivery amount: actTtlGenRT[sid, t]
The value of the power generation day-ahead forecast series at time t: reGenRT[sid,t]
Power suppression amount of variable power supply: curtailRT[sid, t]
Amount of electricity purchased: pchsGen [t]
Battery charging efficiency: esEffc
Battery charge amount: actEsCfRT[sid, t]
Battery charging efficiency: esEffc
Power output of private power generation equipment: actLclFRT [sid, t]

The power generation forecast time series 110 is the output of the high-frequency power generation short-term forecasting means 109. However, constraints such as those in equations (416) to (419) are imposed on the optimization variables related to the first time of the planning target.

Equation (416) shows that any predicted series of battery charge is equal to any other predicted series.

Equation (417) shows that any predicted series of battery discharges is equal to any other predicted series.

Equation (418) shows that any predicted sequence of power curtailment for a variable source is equal to any other predicted sequence.

Equation (418) shows that any forecast series of the power output of the private generating plant is equal to any other forecast series.
Due to these constraints, the variables related to time t, for which the output as a control command must be determined, can be made to take values that are independent of the predicted time series, and it becomes possible to output the command values 112, 113 while maintaining optimality for multiple predicted series, without performing processing such as selecting based on some other criterion.

Third Embodiment
The high-frequency short-term power generation prediction means 109 according to the third embodiment of the present invention includes a neural network that has been machine-learned to learn the relationship between the individual measurement data of the variable power source 151 and the amount of power generated by the variable power source 151. This makes it possible to predict the amount of power generated by the variable power source 151 at a predetermined interval.

FIG. 4 is a diagram showing the configuration of a power generation site 150 where a variable power source 151 is installed.
The high frequency short-term power generation prediction means 109 uses the output of a neural network that has learned the relationship between the captured images 202 of an imaging device 201 installed at the power generation site 150 where the variable power source 151 is installed, and the power generation output information 221 of the variable power source 151. A combination of the power generation output information 221 for the past 30 minutes and the captured images 202 for the past 30 minutes is used as training data, and at the prediction stage, the power generation time series is predicted from the captured images 202 for the past 30 minutes. By using this neural network, it is possible to generate a power generation prediction time series that predicts the power generation amount of the variable power source 151 at a predetermined period.

In addition, instead of learning the power generation output information 221 directly, signal processing such as Fourier transform may be performed to convert it into information on intensity in frequency space, or intensity and phase, and then the relationship with the sky image may be learned. Random number disturbance may be added to the phase and intensity, and multiple predicted time series may be generated by performing an inverse Fourier transform to convert it back into a signal in the time domain.

In addition, a set of light meter data from one or more locations may be used instead of the captured image 202 of the sky. Furthermore, pixel information from one or more specific pixels in the captured image 202 of the sky may be limited and used. In this way, this embodiment can be applied even when there are few computational resources available for implementing a neural network that learns the relationship between the site's weather information and the power generation.

4 is a diagram assuming that the variable power source 151 is a solar power generation system, but a windmill may be used and an anemometer may be used instead of the imaging device 201. According to the third embodiment, control can be performed based on highly accurate power generation predictions according to the site-specific topography, wind conditions, and the like.
Furthermore, by using a device that learns the relationship between such sky images and wind speed data and the power generation amount of the variable power source 151, the present invention can also be applied to power generation methods such as tidal power generation for which no power generation prediction service is provided.

Fourth embodiment
In the first to third embodiments, the day-ahead prediction may be made with a coarse time granularity, and the current-day prediction may be made with the same or finer time granularity as the day-ahead prediction. In this way, a probability-constrained optimal plan using a larger number of prediction sequences can be implemented.

Fifth embodiment
In the first to fourth embodiments, the explanation has been given on the assumption that power supply by private power generation facilities, power generation suppression of variable power sources, and charging and discharging of storage batteries are used as control means. In the current supply and demand adjustment market, buyers of adjustment capacity are limited to grid operators, but in the same way that insufficient power is purchased in the intraday market, a service will be incorporated that secures capacity such as the current tertiary adjustment capacity in advance and activates the required amount during the target period.

To achieve this, the objective function of the day-ahead plan in equation (101) should be modified to equation (501) to create a plan that takes into account the cost of upward market procurement adjustment power pf_xpnsPlsDA[sid,t] and the cost of downward market procurement adjustment power pf_xpnsMnsDA[sid,t].
Equation (501) is an example of a day-ahead planning objective function, where the optimization variables are:

Previous day's bid volume: bidsDA[t0]
Battery charge amount: actEsCfDA [t0]
Discharge amount of storage battery: actEsDfDA [t0]
Upward power procured from the market: pcsFlxPlsDA [t0]
Downward power procured from the market: pcsFlxMnsDA [t0]
Power generation output of private power generation equipment: actLclFDA [t0]
Upward market procurement adjustment amount: actPcsFplsDA [t0]
Downward market procurement adjustment amount: actPcsFmnsDA [t0]

The variables that make up the objective function are as follows:

Daily sales: earningsDA[sid,t0]
Premium: premiumDA[sid,t0]
Cost of downward market procurement adjustment power: pf_xpnsMnsDA[sid,t0]
Cost of upward market procurement adjustment power: pf_xpnsPlsDA[sid,t0]
Operating cost of private power generation equipment: lflexXpnsDA[sid,t0]
Penalty: penaltyDA[sid, t0]

Equations (502) and (503) show constraints on the cost of market procurement adjustment capacity. The variables in equation (502) are as follows:

Cost of downward market procurement adjustment power: pf_xpnsMnsDA[sid,t0]
Base cost in the downward direction: flxSageBasePrice[t0]
Downward power procured from the market: pcsFlxMnsDA [t0]
Downward unit cost: flxSageUnitPrice[t0]
Downward market procurement adjustment amount: actPcsFmnsDA [t0]

The variables in equation (503) are as follows:

Cost of upward market procurement adjustment power: pf_xpnsPlsDA[sid,t0]
Base cost in the upward direction: flxAgeBasePrice[t0]
Upward power procured from the market: pcsFlxPlsDA [t0]
Unit cost in the upward direction: flxAgeUnitPrice[t0]
Upward market procurement adjustment amount: actPcsFplsDA [t0]

The formula (504) is a constraint formula for the delivery amount actTtlGen[sid, t] taking into account the market procurement adjustment power. The variables related to the formula (504) are as follows.

Delivery volume taking into account market procurement adjustment capacity: actTtlGen[sid, t]
Predicted value of market procurement adjustment time series: refVals[sid,t]
Battery charging efficiency: esEffc
Battery charge amount: actEsCfDA [sid, t]
Battery charging efficiency: esEffc
Discharge amount of storage battery: actEsDfDA [sid, t]
Power output of private power generation equipment: actLclFDA [sid, t]
Upward market procurement adjustment amount: actPcsFplsDA [t]
Downward market procurement adjustment amount: actPcsFmnsDA [t]

Equations (505) to (511) are constraints that function the relationship between the procurement amount and the activation amount of the market procurement adjustment capacity. By adding a similar constraint to the daily plan, the market procurement adjustment amount can be reflected in the daily plan.

The formulas (505) and (506) are constraint formulas for the upward market procurement adjustment amount actPcsFplsDA[t]. The variables related to the formulas (505) and (506) are as follows.

Upward market procurement adjustment amount: actPcsFplsDA [t]
Upward direction indicator: pfOplsDA[t]

Equations (507) and (508) are constraint equations for the downward market funding adjustment amount actPcsFmnsDA[t]. The variables related to equations (507) and (508) are as follows.

Downward market procurement adjustment amount: actPcsFmnsDA [t]
Downward direction indicator: pfOmnsDA[t]

From equation (509), the sum of the upward direction indicator pfOplsDA[t] and the downward direction indicator pfOmnsDA[t] is restricted to be equal to or less than 1. The variables related to equation (509) are as follows.

Upward direction indicator: pfOplsDA[t]
Downward direction indicator: pfOmnsDA[t]

From equation (510), the downward power pcsFlxMnsDA[t0] procured from the market is constrained to be equal to or greater than the downward market procurement adjustment amount actPcsFmnsDA[t].

Downward power procured from the market: pcsFlxMnsDA [t0]
Downward market procurement adjustment amount: actPcsFmnsDA [t]

From equation (511), the upward power pcsFlxPlsDA[t0] procured from the market is constrained to be equal to or greater than the upward market procurement adjustment amount actPcsFplsDA[t]. The variables related to equation (511) are as follows.


Upward power procured from the market: pcsFlxPlsDA [t0]
Upward market procurement adjustment amount: actPcsFplsDA [t]

In operation planning, too, by adding the amount of procurement control capacity mobilized as in equation (504) to equation (315) or equation (415), and by adding the cost of procurement control capacity mobilized to the second term on the right-hand side of equation (302) or equation (402), it is possible to create a plan that includes the use of procurement control capacity.

(Modification)
The present invention is not limited to the above-described embodiment, and includes various modified examples. For example, the above-described embodiment has been described in detail to clearly explain the present invention, and is not necessarily limited to those having all of the configurations described. It is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. In addition, it is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.

The above configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware such as an integrated circuit. The above configurations, functions, etc. may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as the programs, tables, and files that realize each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or a recording medium such as a flash memory card or a DVD (Digital Versatile Disk).

In each embodiment, the control lines and information lines are those that are considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are interconnected.

10, 10A Power selling control system 100, 100A Power selling control device 101 Power generation day-ahead prediction means (power generation prediction means)
102 Day-ahead power generation forecast series 103 Day-ahead bid planning means (bid planning means)
104 Bid volume 105 Power generation time-ahead prediction means (power generation prediction means)
106 In-day power generation forecast series 107 In-day bid planning means (bid planning means)
108 Current day buying and selling bid volume 109 Power generation high frequency short-term forecasting means (power generation forecasting means)
110 Power generation forecast time series 111 Operation planning means (resource control series planning means)
112, 113 Command value 114, 115 Contract result 116, 118 Output 117, 119 State 120 Day-ahead market system 130 Intraday market system 141 Demand response business operator 142 Power generation business operator 150, 150a Power generation site 151 Variable power source (an example of a resource)
152 Controllable power source (an example of a resource)
153, 154 Meter 155, 156 Command device 165, 166 Command means 170 Optimization engine 180 Power system 201 Imaging device 202 Captured image 220 Variable power source 221 Power generation output information 222 Variable power source

Claims (12)

a power generation forecasting means for generating a power generation forecast time series that forecasts the power generation amount of a resource for a predetermined period using the neural network and inverse Fourier transform at the bidding stage in the day-ahead market or the same-day market, the power generation output information of the variable power source being a solar power generation facility is Fourier transformed to convert it into intensity information and phase information in a frequency space, and machine- learning the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period;
a bidding planning means for determining a transfer amount that can be provided to the power grid from the power generation amount forecast time series and making a bid;
a resource control sequence planning means for determining a control command sequence for each of the resources that minimizes an error between the contract amount and an actual delivery amount of each of the resources, based on the power generation amount forecast time series, the contract amount determined from the bidding result by the bidding planning means, and the state of each of the resources;
a command means for commanding the resources a part of the control command sequence determined by the resource control sequence planning means that each of the resources can reflect in its control;
A power selling control device comprising: generating a power generation amount prediction time series of the variable power source to be used at each control time during a power transfer period, using the neural network and an inverse Fourier transform;
The power selling control device according to claim 1 . the command means commands the resource, in a period equal to or shorter than the predetermined period predicted by the power generation prediction means, a portion of the control command sequence determined by the resource control sequence planning means that can be reflected in the control of the resource;
The power selling control device according to claim 1 . the bidding planning means determines a transfer amount that can be provided to the power grid from the contract amount determined from the bidding result in the day-ahead market and a time series of the power generation amount forecast of the resource at the bidding stage to the day-ahead market, and bids the transfer amount to the day-ahead market;
The power selling control device according to claim 1 . The bidding planning means determines a commonly available delivery amount on the premise of the operation of different resources.
The power selling control device according to claim 4 . The power generation forecasting means generates at least one power generation forecast sequence in a bidding stage for the day-ahead market or the intraday market.
The power selling control device according to claim 1 . the power generation prediction means generates a power generation prediction series for a predetermined frame in the past and a predetermined frame in the future relative to the current time;
The power selling control device according to claim 1 . The resources include a controllable power source capable of controlling the amount of power generated, and a variable power source whose amount of power generated fluctuates due to an external factor.
The power selling control device according to claim 1 . The variable power source includes any one of a solar power generation facility, a wind power generation facility, and a tidal power generation facility.
The power selling control device according to claim 8 . The resources include any one of a storage battery and a private power generation facility, adjustment power by demand response, and output suppression of a renewable energy power source;
The power selling control device according to claim 1 . a power generation forecasting means for generating a power generation forecast time series that forecasts the power generation amount of a resource for a predetermined period using the neural network and inverse Fourier transform at the bidding stage in the day-ahead market or the same-day market, the power generation output information of the variable power source being a solar power generation facility is Fourier transformed to convert it into intensity information and phase information in a frequency space, and machine- learning the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period;
a bidding planning means for determining a transfer amount that can be provided to the power grid from the power generation amount forecast time series and making a bid;
a resource control sequence planning means for determining a control command sequence for each of the resources that minimizes an error between the contract amount and an actual delivery amount of each of the resources, based on the power generation amount forecast time series, the contract amount determined from the bidding result by the bidding planning means, and the state of each of the resources;
a command means for commanding the resources a part of the control command sequence determined by the resource control sequence planning means that each of the resources can reflect in its control;
A power selling control system comprising: a power generation prediction means performing a Fourier transform on power generation output information of a variable power source, which is a solar power generation facility , to convert the information into intensity information and phase information in a frequency space, and using a neural network that has been machine-trained to understand the relationship between the intensity information and the phase information in the frequency space and a sky image for a predetermined period, and an inverse Fourier transform , to generate a power generation prediction time series that predicts the power generation amount of a resource for a predetermined period at the bidding stage in the day-ahead market or the same-day market;
a bidding planning means for determining a transfer amount that can be provided to the power grid from the power generation amount forecast time series and bidding for the transfer amount;
A step in which a resource control sequence planning means determines a control command sequence for each of the resources that minimizes an error between the contract amount and an actual delivery amount of each of the resources, based on the power generation amount forecast time series, a contract amount determined from a result of bidding by the bidding planning means, and a state of each of the resources;
a step of instructing the resources by a command means regarding a part of the control command sequence determined by the resource control sequence planning means that can be reflected in the control of each of the resources;
A power selling control method comprising:

Images