An Improved Stochastic Gradient Algorithm to Identify PMSM Parameters Based on CAR Models
作者单位：Harbin Institute of Technology Shenzhen Graduate School
会议名称：《第36届中国控制会议》
主办单位：Dalian University of Technology;Systems Engineering Society of China (SESC);Technical Committee on Control Theory (TCCT), Chinese Association of Automation (CAA)
会议日期：2017年
学科分类：080801[工学-电机与电器] 0808[工学-电气工程] 08[工学]
关 键 词：Parameter Identification Convergence Analysis PMSM Stochastic Gradient Algorithm
摘 要：In this paper, we study the parameters identification problem of Permanent Magnet Synchronous Motor(PMSM) in steady state. First, the controlled auto-regressive(CAR) model of PMSM is established. Secondly, based on the obtained CAR model, an improved stochastic gradient algorithm is proposed to identify the electrical parameters of PMSM. By introducing a tuning parameter in the presented algorithm, the current estimation for the unknown PMSM parameters is updated by using the information not only in the current step but also in the previous step. In addition, a convergence result is provided for the developed algorithm. Finally, an example is given to show the advantage of the proposed algorithm for the parameters identification of PMSM.