改进麻雀算法优化Elman神经网络的短期电力负荷预测  

Short-term power load forecasting based on improved sparrow algorithm optimized by Elman neural network

作　　者：邹定江 刘天羽(指导)[1] 王勉 段震宇 ZOU Dingjiang;LIU Tianyu;WANG Mian;DUAN Zhenyu(School of Electrical Engineering,Shanghai Dianji University,Shanghai 201306,China)

机构地区：[1]上海电机学院电气学院,上海201306

出　　处：《上海电机学院学报》2022年第4期221-227,共7页Journal of Shanghai Dianji University

摘　　要：针对埃尔曼(Elman)动态递归神经网络在短期电力负荷预测中容易陷入局部最优、收敛不稳定以及预测精度低的问题,提出了基于Logistic混沌映射初始化种群并加入随机游走扰动的麻雀搜索算法(SSA)优化Elman动态递归神经网络的预测方法。首先,为提高初始解的质量,根据Logistic混沌映射理论对麻雀种群进行初始化;其次,在麻雀搜索食物后,通过随机游走对最优麻雀进行位置扰动,提高其全局与局部搜索能力;最后,将改进麻雀搜索算法(ISSA)与Elman动态递归神经网络相结合,并通过Matlab进行真实数据仿真,对比分析了ISSA-Elman模型与其他模型的电力负荷预测结果。结果表明:本文方法误差更小,预测精度更高。To solve the problems that Elman dynamic recurrent neural network is prone to local optimization,convergence instability and low prediction accuracy in short-term power load prediction,a prediction method is proposed based on the Elman dynamic recurrent neural network optimized by a sparrow search which introduces a Logistic chaotic mapping to initialize population and adds random walk disturbance.First,to improve the quality of the initial solution,the sparrow population is initialized according to the Logistic chaos mapping theory.Second,after the sparrow searches for food,the optimal sparrow is perturbed by random wandering to improve its global and local search capabilities.Finally,the improved sparrow search algorithm(ISSA)is combined with the Elman dynamic recurrent neural network.The simulation is performed based on the real data through Matlab,and the power load prediction results of the ISSA-Elman model and other models are compared and analyzed.The results show that the proposed method can obtain smaller error and higher prediction accuracy.

关 键 词：ELMAN神经网络 负荷预测 LOGISTIC混沌映射 随机游走扰动 麻雀搜索 

分 类 号：TM714[电气工程—电力系统及自动化]