Short-term Load Forecasting Based On The Artificial Neural Network Optimized By CWOA

Accurate short-term load forecasting can ensure normal social life and production,effectively reduce the cost of power generation,and increase economic and social benefits.We know that the accuracy of short-term power load forecasting is related to two aspects.The first is the relevance and comprehensiveness of the selection of influencing factors,but the scientific validity of the prediction methods.Accurate selection of influencing factors and advanced scientific forecasting methods are the main development directions of contemporary short-term power load forecasting.With the continuous development of society,people's living standards are continuously improving,and the factors that affect the power load will surely become more and more.With the continuous improvement of modern technology,the direction of power load forecasting has increasingly focused on advanced forecasting methods.Artificial neural network is one of the main research hotspots.However,too many influencing factors will lead to an increase in the input dimension of the neural network,which will seriously affect the generalizability of the artificial neural network,which will lead to a decrease in the prediction accuracy.Therefore,how to determine the number of appropriate influencing factors to ensure that the input dimension of the neural network will not be too great as to affect the prediction accuracy without losing the information of the influencing factors,should be one of the current short-term load forecasting studies.direction.This article starts from two aspects.One is the analytical modeling of the selection of factors affecting the short-term power load forecasting,and the other is the research on short-term power load forecasting methods.For the analysis of the influence factor selection,this paper mainly uses the factor analysis method and partial autocorrelation function method.First,factor analysis is used to reduce the dimension of the original influencing factors of power load,and the main influence factor with the highest contribution rate is obtained.Then,the targeted partial autocorrelation function analysis was performed on the main influence factors so that the influencing factors were more comprehensive.Factor analysis makes the dimensions of the influencing factors of the input neural network greatly reduced,which ensures the accuracy of neural network prediction under the premise of not losing the information.At the same time,according to the derived main factors to conduct a special partial autocorrelation function analysis,to further enrich the full range of influencing factors,making the choice of influencing factors more targeted.At the same time,chaos theory is introduced into the traditional whale swarm optimization algorithm to improve the convergence speed and optimization ability of the whale swarm optimization algorithm.In this paper,the three major neural networks,BP neural network,least squares support vector machine and extreme learning machine are used to perform short-term power load forecasting.The three neural networks are optimized using the chaos whale swarm algorithm.Finally,the hourly match data of 122 days from March 1,2017 to June 30,2017 in Hengshui City was used as a load forecast sample to conduct an example analysis.The results verified the effectiveness of the improved neural network model.