基于量子粒子群算法优化LSTM的短期风电负荷预测模型  

Short-Term Wind Power Load Forecasting Model Based on Quantum Particle Swarm Algorithm Optimised LSTM

作　　者：李享蔚 郑雅姣 Li Xiangwei;Zheng Yajiao(Liaoning Technical University,Huludao Liaoning 125000,China)

机构地区：[1]辽宁工程技术大学,辽宁葫芦岛125000

出　　处：《现代工业经济和信息化》2024年第7期238-239,242,共3页Modern Industrial Economy and Informationization

摘　　要：为了提高短期电力负荷预测的精度,研究了一种基于QPSO算法对LSTM神经网络进行优化的算法,并根据LSTM神经网络以及QPSO算法的基本原理,利用QPSO算法优化模型隐含层节点数、训练次数和学习率,建立QPSO-LSTM短期风电负荷预测模型[1]。解决了因网络结构及模型参数的不确定性产生的精度问题,并将该模型与传统的神经网络模型进行了对比。仿真结果表明,QPSO-LSTM模型较传统的LSTM模型预测精度更高。In order to improve the accuracy of short-term power load forecasting,an algorithm based on the QPSO algorithm to optimise the LSTM neural network is investigated.According to the basic principles of the LSTM neural network as well as the QPSO algorithm,the QPSO algorithm is utilised to optimise the number of model implicit layer nodes,the number of training times,and the learning rate,and to establish the QPSO-LSTM short-term wind power load forecasting model.The accuracy problem arising from the uncertainty of network structure and model parameters is solved,and the model is compared with the traditional neural network model.The simulation results show that the QPSO-LSTM model has higher prediction accuracy than the traditional LSTM model.

关 键 词：LSTM神经网络 量子粒子群优化算法QPSO QPSO-LSTM 

分 类 号：TM73[电气工程—电力系统及自动化]