The Research Of Current-Loop Bandwith Expanding Technology For AC-Pmsm Servo System

With the development of permanent magnetic materials and AC-Servo system, permanent magnetic synchronous motor (PMSM) servo system, with high-performance and all-digital control strategy, will become a trend of servo control system development inevitably. Therefore, studying on PMSM servo system is important practical significance.In this thesis, firstly, the development of AC-Permanent servo system and conventional Field Oriented Control (FOC) is introduced, and the factors of impacting on the bandwidth of system current loop are analyzed, which in the conventional FOC strategy. To decrease the main delay of servo system, proposed the dual-sampling and dual-updating strategy (DSDU), under the premise condition of without changing the hardware. And, in theory, contrast the system bandwidth between DSDU strategy and FOC strategy under the same PWM cycle, that has proved the DSDU strategy improve the system bandwidth nearly double.To improve the system current loop bandwith furtherly, according to the form of state space function, the model of PMSM in the synchronous rotor frame is derived, and the predictive model of voltage source inverter. And then, the direct predictive control arithmetic based on selecting the optimal switching condition, and a cost function that evaluates the distance of the predicted system state to the desired values are proposed. The PWM predictive control arithmetic based on the dead-beat control, and the limitations and optimal arithmetic are proposed. At last, the thesis contrasts the proposed two predictive control arithmetic in the way of system response, control precision, harmonic content and elapse of time.Finally, the FOC, Direct Predictive Control and PWM Predictive Control simulation model in Matlab/Simulink environment are build up to simulate the proposed algorithm. Also experiment the proposed algorithm in physical experimental setup based on TMS320F2812. Simulation and experimental results verify the correctness of the proposed algorithm.