Fuzzy and PNN-based direct torque control for permanent magnet synchronous motor

This paper investigates an intelligence realization of direct torque control (DTC) in permanent magnet synchronous motor (PMSM) drives. The scheme is based on fuzzy logic con-trol (FLC) and probabilistic neural networks (PNN). The torque error and flux linkage error were all properly fuzzified into several subsets to select a middle state variable accurately, which is the linear function of flux linkage error and torque error. According to the inputs of the middle state variable and flux linkage angle, the proper switching states of the inverter are selected by the designed PNN. Then, a novel PNN and FLC based PMSM-DTC intelligence implementation scheme is presented in the paper. Simulation results show that the proposed strategy makes the torque and flux linkage ripple lower and current smoother dramatically and has a much better steady state performance while keeping a good dynamic performance compared with the basic one.