Torque ripple minimization of PMSM using PI type iterative learning control

Torque ripples due to cogging torque, current measurement errors and non-sinusoidal flux distribution restrict application of PMSM in direct drive system. In this paper, a method based on PI type iterative learning control (ILC) is presented to suppress the ripple. Tuning method of the controller is given in detail, which ensures stability and convergence of the system. Effect of controller parameters on its performance is discussed. Compared with the existing ILC compensatory strategy, the proposed method needs no estimation or close loop control of the electromagnetic torque, which simplifies structure of the controller. Simulating and experimental results verify that ripple of q axis current is reduced obviously, which consequently leads to desired torque ripple restriction performance.