基于权重优化的混合风速预测方法设计  

Design of Hybrid Wind Speed Prediction Method Based on Weight Optimization

作　　者：扈菲宇 黄峰[1] 谢鑫 HU Feiyu;HUANG Feng;XIE Xin(College of Electrical and Information Engineering,Hunan Institute of Engineering,Xiangtan 411104,China)

机构地区：[1]湖南工程学院电气与信息工程学院,湘潭411104

出　　处：《湖南工程学院学报(自然科学版)》2024年第4期9-14,共6页Journal of Hunan Institute of Engineering(Natural Science Edition)

基　　金：湖南省自然科学基金项目(2022J50116、2022JJ50014);湖南省研究生科技创新项目(CX20231290);2022年湘潭市创建国家创新型城市建设专项项目(CG-ZDGS20221004).

摘　　要：风速预测模型各有所长,采用单一模型很难保证预测精度.鉴于此,本文提出一种混合短期风速预测方法,采用长短期记忆网络(LSTM)和时序卷积网络(TCN)进行加权组合得到风速预测结果.首先,采用LSTM和TCN分别对原始风速序列进行预测,分别得到两组预测结果.然后,采用果蝇算法(FOA)寻优,找到预测结果组合的最优权重.最后,将预测结果与其对应的权重系数相乘,再相加得到最终预测结果.仿真结果表明,相比单一预测模型,所提预测方法在RMSE与MAPE指标精度上均得到明显提升.Wind speed prediction models have their own strengths,and it is difficult to guarantee the prediction accuracy by using a single model.A hybrid short-term wind speed prediction method is proposed,which uses a weighted combination of long short-term memory network(LSTM)and temporal convolutional network(TCN)to obtain wind speed prediction results.Firstly,LSTM and TCN are used to predict the original wind speed sequence,and two sets of prediction results are obtained respectively.Then,the fruit fly optimization algorithm(FOA)is used to find the optimal weights for the combination of predicted outcomes.Finally,the prediction results are multiplied by their corresponding weight coefficients and summed to obtain the predictions.Simulation results show that compared with a single prediction model,the proposed prediction method improves the accuracy of RMSE and MAPE indexes by 67.38%and 69.59%respectively.

关 键 词：风速预测 长短期记忆网络 时序卷积网络 果蝇算法 

分 类 号：TM732[电气工程—电力系统及自动化]