Error correction method for weather forecast in short-term load prediction

Abstract

The invention relates to an error correction method for weather forecast in short-term load prediction. The method comprises the following steps: analyzing weather forecast sample data and historical sample data of a day to be predicted in short-term load prediction to obtain Fisher information window data of real weather data 12 hours before each moment to be predicted; obtaining a real meteorological data Fisher information value 12 hours before each to-be-predicted moment on the basis of the window data; carrying out normalization and mapping processing on the calculated Fisher information value according to the characteristics of the Fisher information; obtaining a correction value of weather forecast data at a to-be-predicted moment by using the processed Fisher information value and a weather data change value of one hour before the to-be-predicted moment, and inputting the correction value of the weather forecast data and historical weather and load data into an artificial intelligence model; and correcting the weather forecast data at the moment to be predicted through Fisher information and weather data changes, therefore, the precision of short-term load prediction can be further improved.

Description

technical field [0001] The invention relates to the technical field of power system load forecasting, in particular to an error correction method for weather forecast in short-term load forecasting. Background technique [0002] In short-term load forecasting, improving its forecasting accuracy has always been an important task in load forecasting. Meteorological factors have a huge impact on the accuracy of short-term load forecasting, so meteorological factors will be taken into account in load forecasting. In real load forecasting, the weather forecast data of the time period to be predicted is bound to be used for load forecasting at a certain time in the future. The weather forecast data often has certain errors with the real weather data. Therefore, the weather forecast data Error correction is very meaningful for short-term load forecasting. [0003] The literature "On the Influence of Weather Forecast Errors in Short-Term Load Forecasting Models" "IEEE Transactions ...

AI Technical Summary

Problems solved by technology

[0003] The literature "On the Influence of Weather Forecast Errors in Short-Term Load Forecasting Models" "IEEE Transactions on Power Systems, 2010" proposed the impact of the error between weather forecast data and real weather data on the accuracy of short-term load forecasting, but did not give Take substantive measures to correct errors and improve the accuracy of load forecasting

The document "Short-term load forecasting based on load decomposition and real-time meteorological factors" "Power Grid Technology, 2009" proposed a gray model method to correct the meteorological data, but did not take into account the temperature change before the correction time point

Images