Research On PMSM AC Servo System Based On DSP And Its Neural Netwoks Control Strategies

In this paper, the mathematic model of Permanent Magnet Synchronous Motor (PMSM) is studied. A control system, which includes TI's DSP—TMS320F240 as central control unit, intelligent power module (IPM) as the power switch unit, and PMSM as actuator is designed. Based on it, the control structure that single neuro is used as the outer loop of the speed and the vector control is used as the inner loop of the current is present in this paper and the results verify the scheme above is valid and its dynamic and stationary performance is very good.To enhance the static and dynamic performances, single-neuron adaptive PID control system is introduced in this paper. A single-neuron adaptive PID intelligent controller is obtained by combing neural network with traditional PID regulator. The gain self-amending single-neuron adaptive controller is formed by introducing the approach of recurrence and amendment of gain on-line into the design of single-neuron adaptive PID controller. By detecting the expected value and actual value online, the adaptive control algorithm can be developed,which are not affected by the accuracy of the identified model. This adaptive control algorithm needs not to identify the parameters of the system and the controller has simpler structure and control algorithms and can be realized online easily. The AC servo system based on vector control is designed using Matlab tools. The single neuro speed controller and PMSM model are presented in this paper, and the simulation result shows that its performance is much better than that of PI controller, and the system is much more robust.