基于属性约简与加权最优层次聚类的短期风速混合预测  

Short-term Hybrid Forecasting of Wind Speed Based on Attribute Reduction and Weighting Optimal Hierarchical Clustering

作　　者：秦本双 杨子轶 李琼林 张朔严 张文燕 郭宇[1] QIN Benshuang;YANG Ziyi;LI Qionglin;ZHANG Shuoyan;ZHANG Wenyan;GUO Yu(School of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo 454003,Henan Province,China;Henan Key Laboratory of Intelligent Detection and Control of Coal Mine Equipment,Jiaozuo 454003,Henan Province,China;State Grid Electric Power Research Institute of Henan Electric Power Company,Zhengzhou 450052,Henan Province,China;College of Electrical and Information Engineering,Hunan University,Changsha 410082,Hunan Province,China)

机构地区：[1]河南理工大学电气工程与自动化学院,河南省焦作市454003 [2]河南省煤矿装备智能检测与控制重点实验室,河南省焦作市454003 [3]国网河南省电力有限公司电力科学研究院,河南省郑州市450052 [4]湖南大学电气与信息工程学院,湖南省长沙市410082

出　　处：《电网技术》2024年第5期2054-2063,I0067,共11页Power System Technology

基　　金：国家自然科学基金项目(52277082);河南省科技攻关计划项目(242102241059)。

摘　　要：准确的风速预测是提高风功率预测精度的重要保障。为此,提出一种基于互信息(mutualinformation,MI)属性约简与加权最优层次聚类(weighting optimal hierarchy clustering,WOHC)的离群鲁棒极限学习机(outlier robust extreme learning machine,ORELM)风速混合预测方法。首先,计算32维风速属性特征与风速时间序列间的MI,分析不同特征与风速的相关性。在此基础上,分别采用最大相关最小冗余(maximum correlation minimum redundancy,MRMR)算法和WOHC算法实现风速属性特征的约简及风速样本数据的聚类划分,并通过最优化聚类预处理(clusters optimizationonpreprocessingstage,COPS)确定最优聚类数。然后,采用ORELM对不同样本集分别进行训练,构建ORELM风速混合预测模型。计算待预测点约简后的属性特征与每个聚类中心的欧式距离,选择匹配的ORELM模型进行风速预测。最后,结合东北某风电场实测数据对所提预测方法的有效性和准确性进行验证,结果表明所提方法具有较好的预测精度,能够满足实际风电场风速预测的需要。Accurate wind speed prediction is an important guarantee to improve the accuracy of wind power prediction.Therefore,a hybrid wind speed forecasting method of outlier robust extreme learning machine(ORELM)based on mutual information(MI)attribute reduction and weighted optimal hierarchical clustering(WOHC)is proposed.On this basis,the maximum correlation minimum redundancy(MRMR)and WOHC algorithms are applied to reduce wind speed attributes and cluster the wind speed sample set.The optimal cluster number is determined by the cluster optimization on the preprocessing stage(COPS)method.Then,ORELM is used to train different sample data sets,and the ORELM wind speed hybrid prediction model is established.Calculate the Euclidean distance between the point to be predicted and each cluster center,and select the matching ORELM model to predict the wind speed.Finally,the effectiveness and accuracy of the proposed prediction method are verified by the measured data of a wind farm in Northeast China.The results show that the new method has good prediction accuracy and can meet the needs of wind speed prediction of actual wind farms.

关 键 词：风速 混合预测 属性约简 WOHC ORELM 

分 类 号：TM614[电气工程—电力系统及自动化]