Research On Modeling Of Meteorological Factors And Correction Of Meteorological Forecast In Short-term Load Forecasting

Short term load forecasting(STLF)is an important guarantee for the sustainable development of the entire smart grid,as well as an important aspect of energy management in the power sector and an important guarantee for safe operation.With the steady improvement of the entire social economy and the improvement of people’s quality of life,various types of high-load civilian electrical equipment have emerged one after another.For example,the use of a large number of refrigeration equipment in summer makes the proportion of meteorological load in the total load become larger and larger.The influence of meteorological factors on the load must be considered when carrying out load forecasting.However,in the actual short-term load forecasting,the weather data used is the weather forecast data provided by the weather station.There are errors between the weather forecast data and the actual weather data.Such errors will definitely affect the accuracy of the load forecast.Therefore,the research focus of this article is to deal with the meteorological factors affecting the load and correct the meteorological forecast error,so as to improve the accuracy of STLF.This article is based on the study of the Least Squares Support Vector Machine(LSSVM)parameter model optimized by the improved artificial bee colony(ABC)algorithm and the analysis of the correlation between meteorological factor data and power load data.A new load forecasting method for improving ABC to optimize LSSVM and Fisher Information(FI)processing meteorological data and correcting meteorological forecast errors is proposed.The main research contents are as follows:First,introduce the traditional Support Vector Machine(SVM)and least squares support vector machine models.In order to enhance the performance of the model,the improved ABC algorithm optimizes the penalty parameters and kernel function of LSSVM and establishes a new model.Numerical simulation experiments verify the effectiveness of the new model.Then,by analyzing the correlation between meteorological factors and load data,a weighted processing method for the cumulative effect of the single meteorological factor and the comprehensive meteorological index based on Fisher information is given.At the same time,considering the impact of weather forecast errors on the load forecast accuracy,Fisher information is used to correct the weather forecast data to further improve the forecast accuracy.Finally,the corrected weather forecast data processed by Fisher information are input as input variables into SVM,LSSVM and improved ABC optimized LSSVM models.The simulation experiment results show that the improved ABC optimized LSSVM model based on Fisher information can achieve better forecast accuracy and is more suitable for smart grid weather-sensitive load forecast.