Research On Speed Sensorless Control Of Permanent Magnet Synchronous Motor Against Load Disturbance

At present,permanent magnet synchronous motors(PMSMs)are widely used in high-performance drive systems of motors due to their high efficiency,low loss,small size,and fast dynamic response.Among the high-performance control strategies for PMSM,the vector control strategy is the most widely used and in-depth method.However,in practical applications,PMSMs are subject to various external disturbances such as load disturbance,which will greatly reduce the dynamic performance of the control system.In addition,the speed sensorless control strategy of PMSM can effectively reduce the system size and reduce the cost,but it also needs to improve the robustness to load disturbance and machine parameter perturbations.In this regard,this paper makes a research on speed sensorless vontrol of of PMSM against load disturbance.First,in order to improve the anti-load disturbance capability of the control system,three kinds of load disturbance observers were researched: Longberg observer,extended Kalman filter observer and sliding mode observer.Through analysis,the Romberg observer is simple to implement,but it is greatly affected by mechanical parameters;the extended Kalman filter observer has higher observation accuracy,but it is limited by parameter tuning and complicated calculations;the sliding mode observer has higher robustness to mechanical parameters,but its chattering problem has always limited its application.This paper designs a new type of load disturbance observer based on the second-order sliding mode theory,which has high observation accuracy of load disturbance and has high robustness to mechanical parameters.Then,in order to better combine the load disturbance observer to improve the speed sensorless control performance,this paper chooses the speed sensorless control method based on the flux observer.Firstly,the design of flux observer and its speed estimation method are introduced.However,the current in the flux observer is estimated as open-loop control,which reduces the robustness of the observer to the perturbation of machine parameters.Based on this problem,this paper proposes an improved flux observer based on second-order sliding mode theory,which effectively improves the robustness of the observer to machine parameters,thereby improving the accuracy of speed estimation and torque estimation.Finally,the performance of the proposed various load disturbance observers and flux observers and simulated and experimentally verified.The results verify the feasibility and effectiveness of the proposed method.