Continuous set nonlinear model predictive control for PMSM drives

A continuous set nonlinear model predictive control (CS-NMPC) for a permanent-magnet synchronous machine (PMSM) is proposed. With an effective cost function optimization algorithm, the output voltage vectors are continuous. Compared with finite set model predictive control (FS-MPC), CS-NMPC can decrease the current and torque ripple dramatically, and have the same dynamic performance. Furthermore, it is robust to parameter variations and can successfully decouple the independent control parameters. Using model based design (MBD), CS-NMPC is implemented on a field programmable gate array (FPGA) with parallel and pipeline processing techniques in short execution time. Experimental results illustrate the high performance of the strategy for PMSM drives.