Parameter Observability Analysis of Interior Permanent Magnet Synchronous Motor and Online Parameter Identification Based on Adaptive Extended Kalman Filter

Wei Yin,Xiaofei Chang,Wenwen Lv,Huan Yang,Haibing Wang
DOI: https://doi.org/10.1049/icp.2024.2131
2024-01-01
Abstract:Accurate online identification of motor parameters is of great significance to realize high-performance control of electric drive systems. This paper analyzed the observability of different parameter combinations in steady state and transient state, most of which are rank deficient without persistent excitation. Based on the parameter observability research, this paper selected stator resistance and PM flux linkage as the identification target. It is commonly acknowledged that the effectiveness of extended Kalman filter (EKF) in estimating dynamic states is notably influenced by the covariance matrices associated with process noise ( Q ) and measurement noise ( R ). To address this issue, this paper proposed an adaptive adjustment approach for the covariance matrix, aiming to enhance convergence effect and improve estimation accuracy.
What problem does this paper attempt to address?