基于相似日和IWOA优化BiLSTM的短期电力负荷预测  

Short-term power load forecasting based on similar days and IWOA optimized BiLSTM

作　　者：朱莉[1] 李豪 汪小豪 姜成龙 曹明海 ZHU Li;LI Hao;WANG Xiaohao;JIANG Chenglong;CAO Minghai(Hubei University of Technology,Hubei Key Laboratory for High-efficiency Utilization of Solar Energy and Operation Control of Energy Storage System,Wuhan 430068,China;Hubei University of Technology,School of Electrical and Electronic Engineering,Wuhan 430068,China)

机构地区：[1]湖北工业大学太阳能高效利用及储能运行控制湖北省重点实验室,武汉430068 [2]湖北工业大学电气与电子工程学院,武汉430068

出　　处：《中南民族大学学报(自然科学版)》2025年第4期507-514,共8页Journal of South-Central Minzu University(Natural Science Edition)

基　　金：新能源及电网装备安全监测湖北省工程研究中心开放研究基金资助项目(HBSKF202124)。

摘　　要：为了有效提升短期负荷预测的精度,提出了一种基于相似日和IWOA优化BiLSTM的短期电力负荷预测模型.该模型首先利用Pearson相关性分析选取负荷的主要影响因素,并利用综合匹配相似度选取相似日,为模型提供更有效的输入;然后研究了一种基于非线性控制参数策略和种群变异策略的IWOA算法,对BiLSTM网络的参数进行寻优,构建IWOA-BiLSTM预测模型;最后以澳大利亚真实负荷数据集作为实际算例进行验证,结果表明:该预测模型相较于其他模型获得了更高的预测精度,证明了该方法的有效性.In order to effectively improve the accuracy of short-term load forecasting,a short-term power load forecasting model based on similar days and IWOA optimized BiLSTM was proposed.The model first uses Pearson correlation analysis to select the main influencing factors of the load,and uses the comprehensive matching similarity to select the similar day to provide more effective input for the model;then an IWOA algorithm based on a nonlinear control parameter strategy and a population variation strategy is designed to optimize the parameters of the BiLSTM network and build the IWOA-BilSTM prediction model;finally,taking the Australian real load data set as an example,the experimental results show that the prediction model proposed has higher prediction accuracy than other models,which proves the effectiveness of this method.

关 键 词：短期负荷预测 改进鲸鱼优化算法 相似日 双向长短期记忆网络 超参数寻优 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TM411[自动化与计算机技术—控制科学与工程]