Predictive Flux Control And Periodic Disturbance Rejection For Permanent Magnet Synchronous Machine

Permanent magnet synchronous machines(PMSMs)have been widely applied in aerospace,rail transit,marine propulsion,and servo systems due to its definite advantages of high power density,high energy efficiency,and high control flexibility.With the expanding application of PMSMs,the application conditions are becoming more and more complex.As higher performances are required,the controller design is challenged.As well as highly dynamic performances,enhanced steady-state performances and low torque ripples should be guaranteed in the control of PMSMs.In addition,periodic disturbances caused by current measurement errors and the converter dead-time effect should be attenuated to reduce their influence on steady-state performances and dynamics of the control system.Based on this background,a predictive flux control strategy for PMSM based on extended control set is proposed.The model of PMSM is established in continuous-and discrete-time domains,respectively.With the help of the predictive model,the stator flux trajectories of PMSM are analyzed based on the concept of extended control set.The optimal control voltage vector is selected through sector identification,exhausting evaluation,and magnitude optimization.The duty cycle signals can be directly obtained in the prediction process.Theoretical analysis and experimental results show that,compared with conventional predictive controllers,inaccurate sector determination is avoided as the weighting factor is not required when the proposed algorithm is applied.As a result,the adaptability of the control strategy to different operating conditions can be improved while highly dynamic performances are guaranteed.Meanwhile,the torque and flux ripples can both be effectively suppressed because the variation of flux trajectories within the entire sampling interval is taken into account in the cost function.In addition,periodic disturbances caused by the converter dead-time effect and current measurement errors are analyzed in this paper.On this basis,the periodic disturbance attenuation method based on dual-loop predictive control is proposed.Parameter tuning guidelines of the primary controller and the periodic disturbance attenuation loop are provided,respectively.And the robustness against model mismatch and parameter variations is further analyzed.Theoretical analysis and experimental results show that,compared with conventional periodic disturbance attenuation method,relationship between reference tracking loop and disturbance attenuation loop is released by the proposed controller.The influence of the periodic disturbance attenuation process on the reference tracking process is avoided,which implies the decoupling between reference tracking and disturbance attenuation.As a result,periodic disturbances are strongly attenuated while smooth and fast transient performances are simultaneously guaranteed.