Offline inductance identification of PMSM with adaptive inverter nonlinearity compensation

Offline inductance identification of permanent magnet synchronous motor (PMSM) is essential for self-tuning of closed-loop controller and position observer in sensorless vector controlled drives. This paper proposes an enhanced offline identification method fulfilled at standstill with adaptive inverter nonlinearity compensation for general-purpose drive applications. The d-q axis inductances can be obtained at standstill by using a voltage-source-inverter fed drive. Influences of inverter nonlinearities on direct and quadrature inductance identifications exploiting high frequency (HF) voltage injection at different rotor positions are analyzed. Then an adaptive inverter nonlinearity compensation strategy employing a simple self-learning of compensation coefficients is proposed to enhance the accuracy of inductance identification. The feasibility of the proposed adaptive offline inductance identification method is validated on a 22kW interior PMSM.