Nonlinearity Compensation in Inverters and PMSMs Using an Artificial Neural Network

This paper proposes a compensation method for voltage distortion caused by inverter nonlinearities and distortion induced by the permanent magnet synchronous motors (PMSMs). To mitigate distortions caused by both the inverter and the motor, a compensation voltage was modeled. Since the shape of compensation voltage is difficult to describe theoretically, an Artificial Neural Network (ANN) was used to represent it. The proposed ANN model compensates for the nonlinearities of both the inverter and the PMSMs, resulting in the reduction of 6th-order harmonic current components. The validity of the proposed method is experimentally verified using a PMSM with the back electromotive force (back-EMF) that includes 5th-and 7th-order harmonic components.