基于WOA-VMD-ISSA-LSSVM的商业短期电力负荷预测研究  

Research on Commercial Short-Term Power Load Forecasting Based on WOA-VMD-ISSA-LSSVM

作　　者：杨广亮[1] 万俊杰 YANG Guangliang;WAN Junjie(Jiangsu Acrel Electrical Manufacturing Co.,Ltd.,Jiangyin 214405,China;Acrel Electric Co.,Ltd.,Shanghai 201801,China)

机构地区：[1]江苏安科瑞电器制造有限公司,江苏江阴214405 [2]安科瑞电气股份有限公司,上海201801

出　　处：《现代建筑电气》2023年第8期7-15,共9页Modern Architecture Electric

摘　　要：针对商业中心短期负荷预测难度大、精度低的问题,提出了一种基于WOA-VMD-ISSA-LSSVM的短期电力负荷预测模型。首先,采用鲸鱼优化算法(WOA)对变分模态分解(VMD)的最佳影响参数进行优化选择,得到最佳效果的分解子序列,减少不同趋势信息对预测精度的影响。然后针对麻雀搜索算法(SSA)在迭代后期会出现种群多样性减少和易陷入局部极值的问题,引入Tent混沌序列和动态自适应权重对麻雀搜索算法进行改进,并采用改进的麻雀搜索算法对最小二乘支持向量机(LSSVM)的惩罚参数和核参数进行优化组建ISSALSSVM预测模型。最后将分解后的各组数据分别输入ISSA-LSSVM模型中,并将每个子序列的预测结果相加得到最终预测结果。经某地区商业中心真实电力负荷数据仿真,验证了该模型的有效性,在一定程度上为实现微电网能管理中的发电优化调度、负荷管理提供依据。Aiming at the problems of high difficulty and low precision in short-term load forecasting of commercial center,a short-term power load prediction model based on WOA-VMD-ISSA-LSSVM was proposed.First of all,Whale Optimization Algorithm(WOA)is used to optimize and select the best impact parameters of Variational mode decomposition(VMD)to get the decomposition subsequence with the best effect.Reduce the influence of different trend information on prediction accuracy;Then,aiming at the problem that population diversity is reduced and local extreme value is easy to fall into in the late iteration of Sparrow Search Algorithm,the chaotic sequence and dynamic adaptive weight are introduced to improve sparrow search algorithm.ISSA optimized the penalty parameters and kernel parameters of Least Squares Support Vector Machine(LSSVM)to build the ISSA-LSSVM prediction model.Finally,the decomposed data of each group were input into the ISSA-LSSVM model,and the final prediction result was obtained by adding the prediction results of each subsequence.The simulation of real power load data of a commercial center in a region verifies the validity of the model,which provides a basis for the optimization of power generation scheduling and load management in microgrid energy management to a certain extent.

关 键 词：变分模态分解 鲸鱼优化算法 Tent混沌序列 麻雀搜索算法 最小二乘支持向量机 

分 类 号：TM734[电气工程—电力系统及自动化]