Design of a lying sensor for permanent magnet synchronous machine torque ripple reduction using the iterative learning control technique

Permanent magnet synchronous machines (PMSM) are widely used for high-performance drive systems. However, an important problem for PMSMs is that parasitic torques may degrade the performances of the drive system. These torque ripples generally vary periodically with the rotor position and lead to speed ripple. To suppress these speed ripples, an iterative learning control (ILC) is used, because it is a good candidate for dealing with periodical errors. In this paper, an original approach called “lying sensor technique” is proposed and analyzed. Compared to the torque ripple reduction approaches which realize the current compensation calculation in the controller, this technique consists in modifying the feedback speed information of the sensor. ILC is integrated into this technique for computing the lying speed information. Simulation is used to check the effectiveness of this approach. Simulation results prove that the ILC lying sensor technique has a good performance.