Application Of DPSO-LSSVM Model Based On Dynamic Particle Swarm Optimization In Short-term Power Load Forecastin

Accurate and reliable short-term power load forecasting is not only related to the careful arrangement of modern production,but also the basis and guarantee of scientific and reliable planning of power sector.Due to the increasing proportion of new energy and the increasing requirements of some industries for the accuracy of load forecasting,the current short-term load forecasting methods are difficult to meet the requirements of the new period.We must explore new forecasting methods.The main contents of this paper are as follows:(1)Mastering the basic theory and processing methods of short-term load forecasting can effectively establish the quantitative input of a series of key characteristics of short-term load forecasting,such as load attribute analysis and load cycle characteristics,build a standardized short-term load forecasting process,identify and deal with missing and abnormal values in the process of historical data adoption,and reduce data noise by cleaning samples.(2)Study and master the least squares support vector machine(LSSVM).In view of the decline of prediction performance caused by LSSVM parameters and empirical assignment,particle swarm optimization(PSO)is used for parameter optimization.At the same time,in order to improve the calculation performance of particle swarm optimization algorithm,the PSO parameter is improved to dynamic particle swarm optimization(DPSO),and dpso-lssvm model is constructed to predict the short-term load of regional power grid by integrating DPSO algorithm.(3)Finally,the factors affecting the regional load change in Panjin area and the load characteristics of Panjin power grid are analyzed.Then the dpso-lssvm model,LSSVM model and D5000 system load forecasting platform are used to predict the short-term load of Panjin power grid respectively.The simulation results show that dpso-lssvm can effectively improve the convergence speed and reduce the load forecasting error.The simulation results are analyzed and summarized.In this paper,dpso-lssvm based on dynamic particle swarm optimization algorithm can improve the short-term load forecasting ability of Panjin and effectively improve the accuracy of power load forecasting in Panjin area.This method provides a more diverse method choice for short-term load forecasting of Panjin power grid.