Torque Ripple Minimization and Control of a Permanent Magnet Synchronous Motor Using Multiobjective Extremum Seeking

In this paper, a multiobjective extremum-seeking (MOES) approach is proposed for torque control of a permanent magnet synchronous motor and minimization of its torque ripple. The latter aspect is important in human-machine interface applications such as haptic interfaces requiring smooth torque profiles at slow speeds. The proposed MOES scheme combines an adaptive iterative learning control method with an adaptive proportional-integral (PI) controller, which makes the system less sensitive to load disturbances and improves the control performance for torque regulation during transient events. Experiments are performed on a proof-of-concept exercise machine that generates desired torque profiles and mechanical impedance based on user's preference. The performance of the proposed controller is further compared with a recently proposed adaptive PI controller. The experimental results validate the effectiveness of the proposed controller in terms of torque ripple suppression, steady-state and transient performance, and load disturbance rejection.