摘要：In this paper, based on the combination of particle swarm optimization (PSO) algorithm andneural network (NN), an adaptive neural network internal model control (NNIMC) is designedfor a permanent magnet synchronous motor (PMSM). Firstly, in order to accelerate theconvergent speed and to prevent problems of trapping in local minimum, PSO algorithm isapplied in feedforward neural network to optimize the NN model's and the NN controller’sparameters. For the adaptation of the learning algorithm of the NN controller, gradient descentmethod is used, secondly, to achieve high-performance speed tracking. The robustness andeffectiveness of the proposed PMSM drive scheme is confirmed by simulation tests in theMATLAB/SIMULINK.
