Predictive control of IPMSM using torque sensor feedback for model errors compensation

This paper deals with predictive control of Interior Permanent Magnet Synchronous Motor (IPMSM) using improved mathematical model. The aim is to provide high quality torque control employing simple cost function based on direct torque control and improved mathematical model of the motor which is based on measured nonlinear parameters of the motor. In order to adequately react to model parameter variation or mismatch we use a proportional integral (PI) observer for torque tracking error reduction. The real torque for feedback compensation is measured by the torque sensor. This PI observer sufficiently compensates the IPMSM model parameter mismatch and helps to eliminate the steady state error of the controller. Resulting control requires very little tuning effort and provides high quality control performance producing low torque ripple while achieving highly dynamic torque tracking. Simulations based on real laboratory prototype of the drive of rated power 4.5 kW have been used to test the control performance. An important contribution of this paper is the detailed analysis of the impact of model parameter mismatch to the overall control performance.