JP3316368B2 - Transmission line snow damage alarm system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for predicting whether or not snow damage will occur in a transmission line from weather conditions, and more particularly to a transmission line snow damage which can accurately predict snow damage even if weather forecast data has a forecast error. It relates to an alarm system.

[0002]

2. Description of the Related Art Transmission lines are exposed to harsh natural environments, and are subject to various natural disasters. In particular, during winter snowfall, electrical accidents may occur over a wide area due to snow on transmission lines. In recent years, neural networks (neural networks that mimic the human brain) that classify complex patterns have attracted attention in terms of their pattern classification ability and generalization ability, and are used to predict snow damage. Neural networks have been proposed. That is, if the weather data is input using the weather data in the past snow damage case, the neural network is trained to output the event of “accident” or “no accident”. By inputting weather forecast data into this neural network, it is possible to classify and predict any of the events "accident" or "no accident".

On the other hand, it is also important to quantitatively estimate the amount of snow on the transmission line during snowfall from the viewpoint of preventing accidents.
The estimated value of the amount of snowfall is obtained from weather data such as temperature, wind speed, and precipitation, and the following estimation formula has been proposed as a method thereof. (1) Admirat's equation; "Measurement of snow accumulation model based on examples in Japan and France; 4th International Conference on Icing of Structures by Atmosphere, September 1989" (2) Ervik's Equation; "Experiment Comprehensive measurement model for icing of transmission lines applied to room and field observations; 198
4th International Conference on Atmospheric Icing of Structures, September 2009 ”(3) Makkonen's equation;“ Estimation of wet snow growth on structures; science and technology in cold regions ”(4) Sakamoto Formula: Central Research Institute of Electric Power Industry Also, a system has been devised that outputs a warning when the neural network determines that there is an accident and the estimated snowfall calculated by the above snowfall estimation formula exceeds a predetermined value. ing.

[0004] By the way, the area that is subject to the snow damage warning is
Considering the case of preparing for the occurrence of snow damage or taking measures to prevent snow damage beforehand, it is desirable that the area be a specific limited area. As local weather forecast data, numerical forecast data and real-time data (hereinafter referred to as “three-hour ahead forecast data”) up to three hours ahead of temperature, wind speed, and precipitation in units of 5 km square mesh, , GPV
Numerical forecast data of temperature, wind speed, and precipitation up to 24 hours ahead at the data point (hereinafter, referred to as “24-hour ahead forecast data”) are provided. The three-hour forecast data is announced every hour, while the 24-hour forecast data is announced twice a day, about every 12 hours.

[0005] Considering the time at which a snow damage warning is output, in order to take the above countermeasures, it is better to set the time as early as possible before the occurrence of the accident. More practical than data.

[0006]

However, according to the conventional transmission line snow damage warning system, it is more difficult to forecast as the weather forecast for the earlier time becomes longer.
It is inferior to the time ahead forecast data. Since the prediction accuracy of the weather forecast data directly affects the accuracy of the snow damage warning system using the data, poor prediction accuracy hinders the improvement of the accuracy of the snow damage warning system.

Accordingly, an object of the present invention is to provide a transmission line snow damage warning system capable of performing a high-precision snow damage warning even if the weather forecast data has a prediction error.

[0008]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a power transmission line which predicts snow damage of a power transmission line in advance based on weather data such as temperature, wind speed and precipitation, and issues an alarm. In the snow damage warning system, correction means for correcting weather forecast data for a long time ahead to be warned using weather actual data provided at each time, and predicting snow damage based on the corrected weather forecast data It is configured to provide a prediction unit.

According to this configuration, weather forecast data up to a long time ahead, for example, weather forecast values up to a 24-hour ahead can be obtained on an hourly basis. can get. Therefore, if the accuracy is corrected, a high prediction accuracy can be obtained, and the accuracy of the snow damage prediction can be increased. The correction means uses the actual weather value of each 5 km mesh as the actual weather data provided at each time, and uses the actual weather forecast data up to a long time ahead as the weather forecast data at two grid points at latitude and longitude intervals of 20 minutes.
Weather forecast data up to four hours ahead can be used.

According to this configuration, the absolute value of the weather forecast data up to a long time ahead is corrected by the actual weather data in units of 5 km squares provided every hour, and the weather forecast data up to a long time ahead ( It is possible to obtain higher prediction accuracy than using the weather forecast data up to 24 hours ahead as it is, and to improve the prediction accuracy of the snow damage warning system. Further, the above object is to provide a transmission line snow damage warning system that predicts snow damage of a transmission line in advance based on weather data such as temperature, wind speed, and precipitation, and issues an alarm. A data receiving device that receives the weather data provided at the time, and a weather forecast data correction that corrects the weather forecast data up to the long term ahead of the alarm using the weather data provided at each time A device, a snow damage determination device that determines the presence or absence of snow damage by a neural network based on the correction data from the weather forecast data correction device, a snow accumulation amount estimation device that estimates a snowfall amount from the weather forecast data, And a warning determining means for determining whether or not to issue a warning based on the result of the snow damage determining device and the amount of snow estimated by the snow amount estimating device. It is.

According to this configuration, the weather data acquired by the data receiving device is corrected by the weather forecast data correction device to the weather forecast data for a long time, and the presence or absence of snow damage is determined based on the correction result. The determination is made by the apparatus using a neural network. Based on the weather forecast data and the transmission line parameters given by the weather forecast data correction device, the snow accumulation amount estimation device estimates the snowfall of the transmission line several hours ahead, and calculates the snowfall estimation calculation value and snow damage. Based on the determination result of the determination device, the alarm determination device determines whether the alarm output is appropriate. Therefore, as a result of correcting the absolute value of the weather forecast data long time ahead, the forecast error of the weather forecast data is reduced, and highly accurate snow damage prediction can be performed.

[0012]

FIG. 1 is a block diagram showing the configuration of a transmission line snow damage warning system according to the present invention. Meteorological data (2
The data receiving device 1 is installed to receive the 4-hour ahead forecast data and the 3-hour ahead forecast data, and the weather forecast data correcting device 2 is connected to the data receiving device 1. Further, the weather forecast data correction device 2 is connected to a snow damage determining device 3 and a snowfall amount estimating device 4, and an alarm determining device 5 is connected to these devices to output a warning.

The data receiving apparatus 1 has a precipitation value, a temperature value, a wind speed value, and each of three values for each mesh of 5 km square.
Predicted values up to an hour ahead and hourly predicted values of rainfall, temperature and wind speed up to 24 hours ahead are input. The weather forecast data correction device 2 receives the weather forecast data received by the data receiving device 1 and corrects the 24-hour forecast data based on the mesh actual condition data (three-hour forecast data). The following correction is performed using the actual mesh data of the corresponding mesh at the 24 hour ahead forecast grid point.

First, assuming that the 24-hour ahead forecast data is D (t), the corrected forecast data after t hours is given by equation (1). D _OUT (t) = D _IN (t) + E (t) (1) (where, t is an integer indicating the predicted time of the weather forecast data to be corrected D _IN (t): t-time forecast data before correction E (t) is a correction term of the weather forecast data. Here, E (t) is a correction term, and the mesh real-time data (described as R) and the 24-hour forecast data (D (r ))). This E (t) can be obtained by equation (2).

E (t) = R−D (r) (2) Next, the snow damage determination device 3 determines whether or not snow damage has occurred based on the weather data corrected by the weather forecast data correction device 2. I do. For weather data at the time of the past accident, input a weather signal to the neural network to give a teacher signal indicating that there is an accident. ”And learn by giving a teacher signal. When new weather data is input, it is predicted whether or not an accident will occur three hours ahead, and a prediction of“ with accident ”or“ without accident ”is output. Can be.

The snowfall estimation device 4 has a function of estimating the snowfall based on the weather forecast data provided from the weather forecast data correction device 22, and is corrected by the weather forecast data correction device 2. The weather data is input, and in addition, transmission line parameters such as the diameter and resistance of the transmission line are input. The snowfall estimation device 4 estimates and calculates the snowfall of the transmission line three hours ahead based on the weather data and the transmission line parameters.

Further, the alarm determining device 5 includes a snow damage determining device 3.
It is determined whether or not to output an alarm based on the determination result of the above and the calculated value of the estimated amount of snow accumulation of the snow amount estimation device 4. This alarm determination device 5
Is configured using an AND circuit, and calculates a logical product of whether or not the snow damage determination result of the snow damage determination device 3 and the estimation result of the snow accumulation amount estimation device 4 exceed a preset threshold. Then, if the snow damage determination result is “accident exists” and the snow accumulation amount estimation result is equal to or larger than the threshold, an alarm is generated.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Here, the basis of the optimum condition will be described.
The present inventors compared the 24-hour forecast data with the mesh live data from March 20 to March 27, 1994. The result is FIG. 2, where the horizontal axis represents time and the vertical axis represents air temperature. The solid line in the figure is the temperature forecast value of the 24-hour forecast data at a certain grid point, and the chain line is a certain 5k.
It shows the actual temperature value in m mesh.

From FIG. 2, it can be seen that the change pattern of the temperature forecast value of the 24-hour forecast data is very similar to the change pattern of the actual temperature value on the 5 km mesh. A similar comparison was made for wind speed and precipitation.
It was confirmed that the patterns of change between the 4-hour forecast data and the mesh live data were very similar. However, when looking at the absolute values, there is sometimes a considerable difference between the two. Therefore, by correcting the absolute values of the 24-hour forecast data with the actual values, it is more effective than using the 24-hour forecast data as it is. Highly accurate weather data can be obtained.

[0020]

As is clear from the above, according to the present invention, the weather forecast data up to a long time ahead to be warned is corrected using the weather data provided at each time, and the corrected weather data is corrected. Since the snow damage is predicted based on the forecast data, high prediction accuracy can be obtained, and the accuracy of the snow damage prediction can be increased.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a transmission line snow damage warning system of the present invention.

FIG. 2 is a characteristic diagram showing a result of comparison between 24-hour forecast data and actual mesh data at a certain time;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Data receiving device 2 Weather forecast data correction device 3 Snow damage judgment device 4 Snow accumulation amount estimation device 5 Alarm judgment device

──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Kensuke Saito 1-3-1 Uchisaiwaicho, Chiyoda-ku, Tokyo Inside Tokyo Electric Power Company (72) Inventor Hiroaki Kitagawa 5-1-1 Hidakacho, Hitachi City, Ibaraki Prefecture (72) Inventor Yoshio Ichichi 5-1-1, Hidaka-cho, Hitachi City, Ibaraki Pref. Hitachi, Ltd. Opto-System Research Laboratories (56) References JP-A-7-167965 (JP, A) JP-A-5-134055 (JP, A) JP-A-9-49884 (JP, A) JP-A-6-331756 (JP, A) A) JP-A-8-194070 (JP, A) Toshiyuki Yamamoto, Kimiharu Kanamaru, Hitoshi Kano, Yoshio Ichichi, Yoshifumi Chino, "A method for judging snowfall information on transmission lines using neural networks," IEICE. B, Japan, 1992, Vol. 112, No. 7, p. 585-592 Shuji Tatezaki, Yukichi Sakamoto, Kazuo Mizushima, Seiichi Kasai, Hisae Yamada, Yoshihiro Shimada, "Development of Snow Accretion Forecasting System for Transmission Lines", Report of Central Research Institute of Electric Power Industry, Japan, Central Research Institute of Electric Power Industry, Japan May 1989, T89001, p. 1-31 Toshiyuki Yamamoto, Kimiharu Kanamaru, Hitoshi Kano, Yoshio Ichiji, Yoshifumi Chino, “A method for determining snowfall information on transmission lines using a neural network,” Proceedings of the National Conference of the Institute of Electrical Engineers of Japan, Member of the National Meeting of the Institute of Electrical Engineers of Japan. Association, March 10, 1992, 10-88 to 10-89 (58) Fields studied (Int. Cl. ⁷ , DB name) G01W 1/00-1/18 JICST file (JOIS)

Claims (4)

(57) [Claims] 1. A transmission line snow damage warning system for predicting snow damage of a transmission line in advance based on weather data such as temperature, wind speed, and precipitation, and issuing an alarm, using actual weather data provided at each hour. A transmission line snow damage warning system, comprising: correction means for correcting weather forecast data up to a long time ahead to be warned; and prediction means for predicting snow damage based on the corrected weather forecast data. 2. The method according to claim 1, wherein said correcting means extends to said long time.
To correct the accuracy of the absolute value of
The transmission line snow damage warning system according to claim 1, wherein
M 3. The correction means uses the actual weather value of each 5 km mesh as the actual weather data provided at each time, and uses the GPV ( Grid Point Value) as latitude and longitude as the weather forecast data up to a long time ahead. 3) The transmission line snow damage warning system according to claim 1 or 2 , wherein weather forecast data up to 24 hours ahead at a grid point is used. 4. A transmission line snow damage warning system for predicting snow damage of a transmission line in advance based on weather data such as temperature, wind speed, and precipitation, and issuing an alarm. A data receiving device that receives the provided weather data, and a weather forecast data correction device that corrects the weather forecast data up to a long time ahead to be a warning using the weather data provided at each time. A snow damage determination device that determines the presence or absence of snow damage by a neural network based on correction data from the weather forecast data correction device, a snow accumulation amount estimation device that estimates the amount of snow from the weather forecast data, and the snow damage determination. A transmission line, comprising: a warning determination unit configured to determine whether to issue a warning based on a determination result by the device and a snow amount estimated by the snow amount estimation device. Harm alarm system.