Self-Commissioning of Permanent Magnet Synchronous Machine Drives at Standstill Considering Inverter Nonlinearities

Offline parameter identification of permanent magnet synchronous machines (PMSMs) is essential for proper tuning of the controller and position observer for general-purpose drives with sensorless control. This paper proposes a self-commissioning method of electrical machine parameters at standstill only using a voltage source inverter fed drive. The influence of inverter nonlinearities including the effect of parasitic capacitance, which may cause estimation error, is analyzed. And an error model of inductance identification considering different rotor positions is established. Along with high-frequency sinusoidal signal, a supplementary direct current signal is injected into the estimated direct-axis to attenuate the inductance identification error. In addition, a compensation strategy based on the error model is adopted to enhance the accuracy of inductance identification. For stator resistance identification, the linear regression method is adopted to overcome the influence of inverter nonlinearities by injecting the linearly increasing current signal. The proposed method is promising and robust to extract the resistance information from the gradient coefficient of the voltage variation. The effectiveness of the proposed self-commissioning scheme is validated on a 22-kW PMSM drive.