Research On PMSM AC Servo Control System

The mathematical model of AC PMSM (Permanent Magnet Synchronous Motor) is introduced, and then introduces the principles of SVPWM (Space Vector Pulse Width Modulation) and CHBPWM (Current Hysteresis Band PWM). Simulation models of PMSM AC servo system based on SVPWM and CHBPWM are built in MATLAB/SIMULINK, and simulation results are compared.The permanent magnet synchronous motor is a non-linear, strong-coupling and time-variable system, and during operation it will be disturbed more or less by factors such as load disturbance.PID control algorithm is simple, reliable, so it is widely used in motion control. However, it is limited to achieve the desired control effect in situations such as precise mathematical model of the control object is difficult to be built, or model structure and parameters change in a large range or the control object is complex non-linear system. As a result, fuzzy control becomes a good choice.Two fuzzy control strategies are presented for speed control of PMSM vector control system, and simulation models are built. Compared with conventional PI control, simulation results show that the presented scheme improves stability, control accuracy and response speed of the control system.The neural network has some excellent abilities such as self-learning, nonlinear mapping, and associative memory; also it has fine fault-tolerant capability and robustness. Algorithm based on the neural network has a promising potential use in control systems which have uncertainty and time-variation.Speed loop of PMSM vector control system based on single neuron PI controller and neural network controller are designed to overcome the shortcomings of traditional PI controller in speed loop of PMSM vector control system. Simulation results show that the single neuron PI controller and neural network controller improve stability, control accuracy and response speed of the control system.