Abstract: Permanent magnet synchronous motor has been widely applied in traditional industrial production and the speed control system. However, the motor has the characteristics of nonlinear, strong coupling and multi variable. hich reduces system response ability and anti-interference ability. To ensure the smooth operation of the system, this paper applies RBF neural network identification device to permanent magnet synchronous motor control system, and uses fuzzy logic to optimize the learning step of neural network. As a consequence, it improves the identification precision of the RBF neural network. The simulation results show that he optimized neural network identifier has a good running performance for the speed control of permanent magnet synchronous motor, which has smaller speed overshoot volume and achieves tomooth faster than the traditional PID control.

Key words: permanent magnet synchronous motor, study step length, neural network, fuzzy control