采用测量电压的永磁同步电机多参数在线辨识  

Multiparameter Online Identification for Permanent Magnet Synchronous Machines Using Voltage Measurements

作　　者：吴春 赵宇纬 孙明轩[1] WU Chun;ZHAO Yuwei;SUN Mingxuan(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,Zhejiang Province,China)

机构地区：[1]浙江工业大学信息工程学院,浙江省杭州市310023

出　　处：《中国电机工程学报》2020年第13期4329-4339,共11页Proceedings of the CSEE

基　　金：浙江省自然科学基金项目(LQ18E070005)。

摘　　要：针对永磁同步电机多参数在线辨识存在欠秩、相互耦合、精度低等问题,该文提出一种采用测量电压的多参数在线辨识方法,同时在线辨识d、q轴电感、定子电阻和永磁体磁链4个参数。通过测量电压代替给定电压,消除了逆变器非线性因素对参数辨识的影响,提高参数辨识精度。通过高频正弦电压注入,先辨识d、q轴电感,然后再将电感辨识结果用于电阻和磁链的递推最小二乘辨识。这种分步辨识策略解决了欠秩问题。同时,在d轴注入方波扰动电流信号,进一步解决由于欠秩导致的参数辨识相互耦合的问题。实验结果证明,所提多参数在线辨识算法在加减速、突加突卸负载和不同温度下,均能准确辨识参数。For the multiparameter online identification of permanent magnet synchronous machines(PMSMs),there are problems of under-ranking,mutual coupling and low precision,etc.This paper proposed a multiparameter online identification method based on voltage measurements,which could identify d-and q-axis inductances,stator resistance and permanent magnet(PM)flux linkage simultaneously.By using the measured voltages instead of the reference voltages,the impact of inverter nonlinearities on the parameter identification was eliminated,and the parameter identification accuracy was improved.First,through high-frequency(HF)sinusoidal voltage injection,the d-and q-axis inductances were identified.Then,the inductance identification results were used for stator resistance and PM flux linkage identification using the recursive least squares algorithm.This separate identification strategy could solve the under-ranking problem.Meanwhile,the low-frequency(LF)square-wave perturbation current signal was injected into the d-axis to further solve the problem of mutual coupling of parameter identification.The experimental results verify that the proposed multiparameter online identification algorithm can identify the parameters accurately under step speed,step load and different temperatures.

关 键 词：永磁同步电机 在线参数辨识 测量电压 递归最小二乘算法 

分 类 号：TM35[电气工程—电机]