Research On Current Sensorless Model Predictive Control For Permanent Magnet Synchronous Motor Systems

Permanent magnet synchronous motor(PMSM) has been widely used in industry for its significant advantages, such as small size, simple structure, fast dynamic response speed, high torque power density and low lose. Vector control(VC) and direct torque control(DTC) are mainly used in the control method of PMSM drive system. DTC system has a faster torque response than VC, but its torque flux ripple is larger. Thus, model predictive control(MPC) was used in this thesis to study the control of PMSM drive system, which not only has fast torque response, but also can effectively reduce torque & flux ripple.For high precision control of PMSM drive system is achieved, closed-loop current control is generally required, which needs some current sensors to measure current information. However, too many current sensors are applied not only to increase costs, but also to increase the complexity of the system and to reduce reliability. A few scholars study the problems about current sensors malfunction in PMSM drive system. This thesis did relatively research on two cases including one phase current sensor normal and current sensorless on the basis of the MPC system. The main contents are as follows.(1) PMSM MPC system is designed by using the method of MPC. Through the comparison and analysis with DTC system, demonstrate the superiority of MPC system compared to the DTC system, and from the angle of the lower switching frequency and harmonic distortion of the MPC system has made the improvement.(2) When one phase current sensor works, using the DC bus voltage, one phase current and position angle information, PMSM model predictive control system was designed based on adaptive observer. The designed adaptive observer can estimate another phase current accurately and identify stator resistance online. The stator currents and stator resistance adaptive observer estimated is applied to the MPC system, achieved better control effect.(3) In the current sensorless circumstances, by using the DC bus voltage, inverter switching status and speed information, PMSM model predictive control system was designed based on adaptive backstepping observer. The designed adaptive backstepping observer can estimate stator currents and stator resistance online at the same time. The stator currents and stator resistance adaptive backstepping observer estimated is applied to the MPC system, achieved the ideal control effect.The designed observer by using the two methods above can not only identify stator resistance online, but also improve stator resistance robustness of observer and enhance stator resistance robustness of model predictive controller. Finally, simulation experiments were taken to verify the feasibility and effectiveness of the proposed method.