基于降维聚类的双馈风力发电机参数辨识  

Parameter identification of doubly-fed wind turbine based on dimensionality reduction clustering

作　　者：吴林林 张家安 刘东 李飞 王潇 刘辉 Wu Linlin;Zhang Jiaan;Liu Dong;Li Fei;Wang Xiao;Liu Hui(State Grid Jibei Electric Power Co.,Ltd.Research Institute,North China Electric Power Research Institute Co.,Ltd.,Beijing 100045,China;College of Artificial Intelligence and Data Science,Hebei University of Technology,Tianjin 300401,China;Department of Information Management,Hebei University of Architecture,Zhangjiakou 075400,China)

机构地区：[1]国网冀北电力有限公司电力科学研究院,北京100045 [2]河北工业大学人工智能与数据科学学院,天津300401 [3]河北建筑工程学院信息管理系,河北张家口075400

出　　处：《可再生能源》2021年第12期1635-1640,共6页Renewable Energy Resources

基　　金：河北省自然科学基金创新群体项目(E2020202142)。

摘　　要：电力系统在调度运行中,须要对风电场风机控制器参数进行辨识。提出了一种基于降维聚类的双馈风机参数辨识方法。首先进行控制器状态方程进行差分线性化,对相关采集数据应用LLE算法进行降维;然后应用K-means聚类进行数据提取;最后应用改进线性神经网络实现参数辨识。其中,LLE算法使数据局部线性特征得以保存,K-means聚类通过提取部分线性数据去除各种扰动和非线性数据,改进的多学习率线性神经网络确保了参数的精确辨识。通过实际双馈风力发电机运行数据作为算例,验证了算法的有效性,降低了参数辨识误差。In the dispatching operation of power system,it is necessary to identify the parameters of wind turbine controller in wind farm.A parameter identification method of doubly fed wind turbine based on dimension reduction clustering is proposed.Firstly,the state equation of the controller is linearly differentiated,and LLE algorithm is used to reduce the dimension of the collected operation data.Then,K-means clustering is used for data extraction.Finally,the improved linear neural network is applied to realize parameter identification.In the algorithm,LLE algorithm retains the local linear characteristics of data.The linear part of the data is extracted by K-means clustering,and the disturbed and nonlinear data are removed.The improved multi learning rate linear neural network improves the accuracy of parameter identification.Taking the actual operation data of doubly fed wind turbine as an example,the effectiveness of the algorithm is verified and the parameter identification error is reduced.

关 键 词：双馈风力发电机 参数辨识 LLE降维 K-MEANS聚类 线性神经网络 

分 类 号：TK81[动力工程及工程热物理—流体机械及工程]