Research On Sensorless Control System Of PMSM For Electric Vehicles

In the background of energy conservation and environment protection, the electric vehicles(EV) have become an important direction for automobile development in the future. Perma-nent magnet synchronous motor(PMSM) is wildly applied in EV drive systems for its nu-merous advantages. However, the reliability, stability and cost of the control system are chal-lenged because of the mechanical sensor. On the other hand, the overall performance of the control system depends mostly on that of the designed controller. From these two aspects, sensorless control technology and controller design are significant in both theoretical and en-gineering ways, which are also hotspots in PMSM control technology research area.In this paper, PMSM control system for EV is set as the research object. Sensorless control technology and controller design are respectively researched to improve the overall perfor-mance of the control system.The mathematical model of PMSM and the space-voltage vector pulse width modulation technology(SVPWM) are initially introduced in the paper. On these basics, in-depth research on vector control theory is put forward, and then the overall control scheme is established. The outstanding performance is verified by the simulation experiment.The sensorless control technology based on model reference adaptive system(MRAS) is researched to solve the mechanical-sensor-related problems. The speed identification scheme is proposed based on the principles of MRAS. The adjustable model and reference model are respectively established. Also the adaption law is calculated with its stability proof. The fea-sibility of the speed identification scheme is verified through simulation experiment.The controller design is put forward for further improvement on the dynamic and static performance of the control system. The predictive current controller is designed to replace the traditional PI controller in the current loop. Considering that the performance is much related to the précised parameters of the model, a disturbance observer is designed to compensate the inaccuracies. Aiming at the serious integral saturation problem in the speed loop, an an-ti-windup PI controller is proposed to refine the step response. The simulation experiment ve-rified the excellent performance of the designed controller.Finally, the experimental platform based on Expert 3 is built with the software design in PE-View environment completed. Through the repeated debugging and numerous experi-ments for the control system on the platform, the speed identification scheme based on MRAS is proved feasible and the controller design is verified effective. At last, the designed control system met the expectations for a more satisfactory dynamic and static performance.