Simulation And Comparative Research On Permanent Magnet Synchronous Servo Motor Control Strategies

Servo motors in the industrial field of mechanical elements used to controloperation can be achieved accurately speed and position control. Ac servo motorbecomes the mainstream of the servo system instead of the dc motor. Thepermanent magnet synchronous servo motor (PMSM) has been widely appliedbecause of its superior performance.The control strategy of PMSM impact on the performance of the servosystem directly. The control to PMSM has always been a research focus. Butbecause of the actual servo system with uncertainties, external disturbance andinternal disturbance, general PID control is difficult to meet the controlrequirements. Therefore the neural network control, sliding mode control ofvariable structure and other control strategies is proposed to better improve thework performance of the system.The PMSM mathematical model is established in this paper. Theconventional PID control has been discussed, and incremental PID and integralseparation have been combined. Moreover, with the learning ability of neuralnetwork, the parameters of PID controller can be adjusted by increasing neuralnetwork which can improve the performance of the PID controller. Sliding modecontrol is superior to conventional PID control on the control performance,sliding mode control can make the system work in the sliding mode surface stably,thus can improve the stability of the system. Since the Cerebella modelarticulation controller (CMAC) have fast learning ability. In order to make thesystem more stable, construct the sliding mode controller based on CMACnetwork, and use the CMAC network to compensate the sliding mode control, toachieve the purpose of speed up the convergence rate. The result of variouscontrol strategies are got after simulation calculation that the neural network canimprove the performance of PID controllers and sliding mode controller. Do some comparison and simulation analysis on the sliding mode control based onCMAC network, conventional PID control algorithm, neural network PID controland sliding mode control. It can be seen that the sliding mode control systembased on CMAC network can get improved and strengthened in both convergencespeed and tracking accuracy.