电力系统保护与控制2024，Vol.52Issue(14)：95-102,8.DOI:10.19783/j.cnki.pspc.231431
基于多策略改进金豺算法优化LSTM的短期电力负荷预测
Short-term power load forecasting based on multi-strategy improved golden jackal algorithm-optimized LSTM
摘要
Abstract
There are problems of low accuracy and poor stability in short-term load forecasting using long short-term memory(LSTM)neural networks.Thus this paper proposes an improved golden jackal optimization(IGJO)algorithm to optimize the LSTM model.First,it integrates a convex lens reverse learning strategy for better starting positions.It introduces the sigmoid function to change the escape energy and balance exploration and development stage.It fuses whale optimization algorithm's spiral enclosure to improve exploration capability and convergence accuracy.Then,it introduces the LSTM neural network,and uses the IGJO algorithm to optimize its hyperparameters and to establish the IGJO-LSTM short-term electricity load forecasting model.Finally,the IGJO-LSTM short-term load forecasting model is validated using actual power load data from a region in Henan province.The experimental results show that the short-term load prediction results of the IGJO-LSTM model at different times on weekdays and weekends are closer to the actual load.Compared to traditional methods,it demonstrates higher accuracy and stability,indicating practical application potential.关键词
电力负荷预测/长短时记忆网络/凸透镜成像/非线性逃逸能量/螺旋包围机制Key words
electricity load forecasting/long-short-term memory networks/imaging of a convex lens/nonlinear escape energy/spiral envelope mechanism引用本文复制引用
王延峰,曹育晗,孙军伟..基于多策略改进金豺算法优化LSTM的短期电力负荷预测[J].电力系统保护与控制,2024,52(14):95-102,8.基金项目
This work is supported by the National Natural Science Foundation of China(No.62272424 and No.62276239).国家自然科学基金项目资助(62272424,62276239) （No.62272424 and No.62276239）
国网河南省电力公司科技项目资助(5217S0240001) （5217S0240001）
