PMSM Periodic Torque/Speed Ripple Reduction Strategy

AC servo system is widely used in industry and motion control field because of its superior performance.In common servo systems,there is a wide range of torque/speed fluctuations caused by non-ideal factors such as current sensor scaling errors,inverter nonlinear factors,and large speed delays.These periodic harmonics can seriously degrade the performance of the serv o system.At the same time,if the speed error is large,performance of the algorithms such as fault detection,load torque observation,and moment of inertia ide ntification will be seriously affected.Therefore,it is of great practical significance to suppress the periodical torque/rotational speed fluctuation of the electromechanical system.This paper first establishes a mathematical model for each harmonic cau se of the torque/speed ripple.Secondly,the corresponding suppression algorithm is adopted for different positions of harmonics in the control loop.In this paper,the notch filter method and Proportional-Integral-Resonant Controller are mainly used.Simulation and experiment verify the effectiveness of the suppression algorithm.The experimental results show that the PIRC has the best obvious suppression effect on the pulsating speed when it is used in parallel,but it has the problem that the higher harmonic amplitude increases.For the problem of higher harmonic content occurring after using PIRC,this paper explores its causes: the inaccuracy of the speed measurement process.To address this problem,the Kalman filter(KF)is used in this paper to impro ve the accuracy of the speed measurement algorithm.It has two main difficulties when it is practically implemented: the parameters of the Kalman filter is difficult to select,and the knowledge of the moment of the inertia in the motor mathematical model is difficult to obtain accurately.For the problem that the parameters of the Kalman filter are difficult to select,this paper deduces and designs an adaptive law to make the parameters automatically adjusted.At the same time,the problem that it is difficult to accurately acquire the knowledge of the moment of the inertia.In this paper,the recursive least squares(RLS)method is used to obtain the required information of the inertia in KF in real time.Both the simulation and the experiment verify the effectiveness of the algorithm.Finally,the above algorithm is verified via the experiment.The notch filter method has a poor suppression effect.Meanwhile,the PIRC used in parallel has the best suppression effect.The phenomenon that high harmonic content increases when implemented with several PIRC can be obviously suppressed by adopting an adaptive Kalman filter(AKF).