基于MIC-VMD-GWO-LSTM的短期风电功率预测  

Short-term wind power prediction based on MIC-VMD-GWO-LSTM

作　　者：梁昌侯 龙华[1] 李帅 周筝 严北斗 LIANG Changhou;LONG Hua;LI Shuai;ZHOU Zheng;YAN Beidou(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China;Datang Yunnan Power Generation Co.,Ltd.,Kunming 650011,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500 [2]大唐云南发电有限公司,云南昆明650011

出　　处：《现代电子技术》2023年第22期115-120,共6页Modern Electronics Technique

基　　金：国家自然科学基金项目(61761025)。

摘　　要：准确的风电功率预测可以合理安排风电场的发电计划和提高电网稳定性。针对单一预测模型预测精度低的问题,提出一种基于MIC-VMD-GWO-LSTM的短期风电功率预测模型。首先使用最大互信息系数法(MIC)对高维特征的风电数据集进行特征提取,以降低数据复杂度;然后采用变分模态分解(VMD)技术将风电功率序列分解为不同频率的模态,以减少功率数据的波动性;接着对每个模态建立GWO-LSTM预测模型,并利用灰狼优化(GWO)算法LSTM模型的相关参数进行优化;最后将每个模态的预测结果求和重构,得到最终的预测结果。仿真结果表明,相对于单一的BP和LSTM预测模型,基于MIC-VMD-GWO-LSTM的组合预测模型的MAPE分别降低了43.16%和31.14%,可有效提高预测精度,证明了该方法在风电功率预测运用中的有效性和可行性。The accurate wind power prediction can reasonably arrange the power generation plan of wind farms and improve the stability of power grid.In allusion to the problem of low prediction accuracy of single prediction model,a short-term wind power prediction model based on MIC-VMD-GWO-LSTM is proposed.The maximum mutual information coefficient(MIC)method is used to extract the features of high-dimensional wind power data sets to reduce the data complexity.The variational mode decomposition(VMD)technique is used to decompose the wind power series into modes of different frequencies to reduce the fluctuation of power data.The GWO-LSTM prediction model is used for each mode,and the relevant parameters of the LSTM model are optimized by means of the grey wolf optimization(GWO)algorithm.The prediction results of each mode are summed and reconstructed to obtain the final prediction results.The simulation results show that in comparison with the single BP and LSTM prediction models,the combined prediction model based on MIC-VMD-GWO-LSTM has a reduced MAPE of 43.16%and 31.14%,respectively,which can effectively improve prediction accuracy.It proves the effectiveness and feasibility of this method in the application of wind power prediction.

关 键 词：短期风电功率预测 最大互信息系数 变分模态分解 灰狼优化算法 长短期记忆 风电功率序列 

分 类 号：TN919-34[电子电信—通信与信息系统] TM614[电子电信—信息与通信工程]