PSO with Exploring Behavior Algorithm and Its Application in Prediction of Power and Economic Output Value

LUO Ji1，WANG Guo-rui1，TANG Ruo-li2，FANG Yan-jun2，ZENG Zhi-jian1，QIAO Jia-geng1
（1.Guangdong Power Grid Company，Guangzhou 510050，China；2.Department of Automation，Wuhan University，Wuhan 430072，China）
Abstract：According to the characteristics of the province industry economy based on industry classification，counting and analyzing the data of historical power consumption and industrial gross production value，and building the prediction model based on power consumption analysis. Proposing a novel improved-PSO algorithm which has a better performance of high complexity function optimization，and applying it to parameter optimization of the prediction model. Result of simulation shows that the built prediction model and improved PSO-based  prediction algorithm can predict the industrial gross production value effectively and reliably.
Key words：power consumption analysis；industrial gross production value prediction；particle swarm optimization（PSO）；neural network