Short-Term Load Forecasting Model of Ameliorated CNN Based on Adaptive Mutation Fruit Fly Optimization Algorithm

作     者：Sun, Kai Dou, Zhenhai Zhang, Bo Zou, Hao Li, Shengtao Zhu, Yaling Liao, Qingling 

作者机构：Shandong Univ Technol Sch Elect & Elect Engn Zibo Shandong Peoples R China Zibo Power Supply Co State Grid Corp China Zibo Shandong Peoples R China 

出 版 物：《ELECTRIC POWER COMPONENTS AND SYSTEMS》 (Electr. Power Comp. Syst.)

年 卷 期：2022年第50卷第1-2期

页      面：1-10页

核心收录：

学科分类：080801[工学-电机与电器] 0808[工学-电气工程] 08[工学] 

基　　金：National Key Research and Development Program of China [2017YFB0902800] 

主　　题：short-term load forecasting convolutional neural network extreme learning machine adaptive mutation fruit fly optimization algorithm 

摘      要：In order to improve the accuracy and calculating speed of load forecasting for the strong nonlinear problem of short-term load, this article proposes a Short-term Load Forecasting Model of Ameliorated CNN Based on Adaptive Mutation Fruit Fly Optimization Algorithm. This method integrates the Extreme Learning Machine (ELM) algorithm into the Convolutional Neural Network (CNN): replace the fully connected layer in the original CNN network with ELM to form a CNN-ELM network. The purpose is to improve the calculation accuracy. An Adaptive Mutation Fruit Fly Optimization Algorithm (AMFOA) was proposed to reduce the probability that the Fruit Fly Optimization Algorithm (FOA) would easily fall into a local optimal value. And then AMFOA is used to optimize the parameters in CNN-ELM network. The above model is used to predict the grid load of a certain area in northern China. Compared with other prediction algorithms, it is proved that the model proposed in this article has higher prediction accuracy and also proved that the model has higher calculation speed than other models.