Model Predictive Control for Torque Ripple Suppression Caused by Misalignment of the Gearbox

Abstract

This paper presents a model predictive control (MPC) based harmonic current injection method (HCIM) to suppress the torque ripple caused by misalignment in gearbox. As it is inefficient to determine the amplitude and phase of harmonic currents manually, the MPC is used to search the appropriate harmonic currents to implement HCIM efficiently. A cost function for MPC is designed to minimize a resultant torque ripple which is calculated by summing the misalignment torque ripple and compensation torque. The compensation torque is an input variable of the predictive model. In order to predict the resultant torque ripple accurately, an estimation method of the misalignment torque ripple with torque sensor is also proposed. The calculated compensation torque is converted into harmonic currents and the harmonic currents can be injected into a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">q</i> -axis current. As a result, the misalignment torque ripple can be suppressed. Experimental results demonstrate verification for estimation method of misalignment torque ripple, and effectiveness of the HCIM. In addition, the proposed MPC-HCIM is also verified by experiments, and the misalignment torque ripple is effectively suppressed by using the proposed method. By using the proposed method, high degree of precision in torque control can be achieved with low quality gearbox which has misalignment.