ILC-Based Voltage Compensation Method for PMSM Sensorless Control Considering Inverter Nonlinearity and Sampling Current DC Bias

Inverter nonlinearity and sampling current dc bias can cause harmonic distortion of estimated back electromotive forces (EMFs), and therefore results in position estimation error and degrades control performance of model-based sensorless techniques of permanent-magnet synchronous motors, especially in low speed region. In this paper, the effects of inverter nonlinearity and sampling current dc bias on back EMF estimation are analyzed. A voltage feedforward compensation method based on an iterative learning control (ILC) algorithm is proposed and it is used to eliminate the harmonics of estimated back EMFs and position estimation error, so as to improve the performance of sensorless control. Compensated voltages at different rotor positions are calculated online by the ILC algorithm and stored in a lookup table. The method does not depend on current polarity and inverter model. And it is simple and easy to implement. The effectiveness of the proposed compensation method is verified by experiments.