Novel Lexicographic MPC for Loss Optimized Torque Control of Nonlinear PMSM

Abstract

In electrical drive applications torque tracking is conflicting with the subsidiary goal of loss minimization. In this Paper a lexicographic optimization is proposed to solve this conflict by stringent prioritization. A novel model predictive control (MPC) with a single objective function is presented which is proofed to be equal to lexicographic optimization in steady state. The loss weighting factor does not affect the steady state tracking offset and no current set points are needed. Results are confirmed by a simulation with an experimental validated machine model. The torque of an anisotropic permanent magnet synchronous machine (PMSM) fed by a voltage source inverter (VSI) is controlled with minimal ohmic losses. The optimization under input constraints is done using the projected fast gradient method (PFGM).