基于相关熵的光伏电池模型鲁棒参数辨识法  

Correntropy Based Robust Parameter Identification for Photovoltaic Cell Model

作　　者：陈倩[1] 张正江[1] 胡桂廷 郑崇伟[1] 闫正兵[1] 朱志亮[1] CHEN Qian;ZHANG Zheng-jiang;HU Gui-ting;ZHENG Chong-weL YAN Zheng-bing;ZHU Zhi-liang(National-Local Joint Engineering Laboratory of Electrical Digital Design Technology,Wenzhou University, Wenzhou 325035,China)

机构地区：[1]温州大学电气数字化设计技术国家地方联合工程实验室,浙江温州325035

出　　处：《控制工程》2019年第1期137-143,共7页Control Engineering of China

基　　金：国家自然科学基金项目(61703309);浙江省自然科学基金项目(LY18F030014);浙江省科技计划项目(2015C31157;LGG18F010016);浙江省大学生科技创新活动计划暨新苗人才计划(2017R426019)

摘　　要：参数辨识可以提高光伏电池模型参数设置的精确度,得到与实际对象相一致的模型。针对加权最小二乘参数辨识(WLS-PI)方法受测量数据显著误差影响显著的问题,提出了一种基于相关熵的鲁棒参数辨识(C-PI)方法。首先,分析了太阳能电池理论模型;其次,构造了一种相关熵估计器,通过该估计器的影响函数分析测量误差对辨识结果的影响,进一步将C-PI方法应用于光伏电池模型的参数辨识上。最后,将C-PI和WLS-PI两种方法用于仿真实例中,结果显示C-PI方法鲁棒性更好;进一步采用光伏电池的测试数据验证了C-PI方法的可行性与有效性。Parameter identification can improve the accuracy of model parameters in the photovoltaic cell model,which can guarantee that the identified model is consistent with the actual plant.In this paper,considering the problem of the weighted least square based parameter identification(WLS-PI) method that the gross measurement errors severely bias the results of identification,a robust parameter identification method is proposed.First,the theoretical model of the photovoltaic cell is investigated.Secondly,the robustness of proposed correntropy estimator is analyzed by using the influence function,and then a correntropy based parameter identification(C-PI) method is developed for the photovoltaic cell model.Finally,the WLS-PI and C-PI methods are used in the simulation example.The results show that the C-PI method is more robust than WLS-PI method.The experimental data of the photovoltaic cell is also used to demonstrate the feasibility and effectiveness of C-PI method.

关 键 词：光伏电池模型 相关熵 加权最小二乘 参数辨识 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]