A Robust Sensorless Full Parameters Identification Algorithm for Permanent Magnet Synchronous Machines

Accurate tuning of vector controlled Permanent Magnet Synchronous Machines (PMSMs) is based on the precise information of their electrical model parameters. This article proposes a novel and robust parameters identification approach for PMSMs. The proposed method involves four successive steps for identifying four main parameters of the equivalent d-q axes model of PMSMs, which are stator resistance, d-q axes inductances and linkage flux. Despite most of conventional parameters identification methods for PMSMs, all of these steps are realized free of position or speed sensors. The first three steps that are respectively assigned to stator resistance and d and q axes inductances identification, are fulfilled in the standstill condition. In the last step, a simple open-loop V/f control system is implemented for identifying the flux linkage value. This method is applicable for both types of PMSMs: Surface mounted PMSMs (SPMSMs) and Interior mounted PMSMs (IPMSMs). Clearly, as the d and q inductance are equal in case of a SPMSM, the third step will be skipped in their parameter identification process. The functionality of the proposed method has been evaluated through some simulations for both types of PMSMs in MATLAB/Simulink environment. Simulation results for PMSMs with various types and different power ratings, confirmed that the proposed method is able to identify all d-q axes model parameters with high accuracy. Accordingly, it can be served as an auto-tuning phase in industrial PMSM drive systems.