Implementation of Continuous Control Set Model Predictive Control Method for PMSM on FPGA

The generalized predictive control (GPC-MPC) algorithm adopts an optimal control strategy, which requires online matrix inversion operation and is difficult to be applied to real-time control systems. Therefore, the permanent magnet synchronous motor (PMSM) drive experiments, using the GPC-MPC algorithm with constraints, are primarily performed in a simulation environment. This paper proposes a new matrix inversion and its circuit design method to complete the above experiments in an embedded environment. By using the calculation steps of matrix block decomposition and reinversion, the parallelism of matrix calculation and the regularity of storage address are improved, and the speed of matrix inversion is accelerated. Also, the model-based design (MBD) method is used to complete the design and verification of the rest of the algorithm, which speeds up the implementation and deployment of the algorithm. Finally, the GPC-MPC algorithm control experiments with current constraints are implemented on the Field programmable gate array (FPGA) experimental board. The experimental results show that the proposed design method has a good control effect and computational efficiency.