Inductance–Resistance Online Identification for Sensorless High-Speed PMSM Considering Resistive Characteristic of Invertor Nonlinearity
Yinfeng Hu, Wei Hua, Mingjin Hu, Yuchen Wang
Abstract
Parameter identification is becoming an essential technology in machine driven industrial drives due to the requirements of high-performance control and online monitoring. However, for high-speed permanent magnet synchronous machines (HSPMSMs), due to the complexity and couplings between discrete-time model and low-value resistance, multi-parameter estimation faces challenges. This paper proposes a non-iterative method to simultaneously obtain resistance and inductance of HSPMSM based on multi-frequency disturbance injection (MFDI). By injecting disturbances with different frequencies, the identification accuracy can be guaranteed not affected from the estimated rotor position error of sensorless control and flux linkage error. Further, it is revealed that the voltage distortion caused by the dead-time effect (DTE) could affect as a resistance, which is modeled as an equivalent dead-time effect resistance (EDTER) in this paper, and consequently, the brought errors for resistance identification are suppressed. In addition, the least square method is employed to solve the result fluctuation problem. At last, a noninvasive temperature estimation of stator windings is accomplished with a precision within ±10°C. Finally, experiments are carried out on a 60000r/min HSPMSM to verify the proposed identification method.