Research On Strong Robust Predictive Control Of Permanent Magnet Synchronous Motor

With the research of motor control technology in recent years,Permanent Magnet Synchronous Motor(PMSM)has been studied by researchers in various countries because of its high power density,small moment of inertia,and flexible design of size structure.PMSM is widely used in petrochemical,aerospace and other fields.It is widely used in petrochemical industry,Numerical control machine,aerospace and other fields.The conventional control algorithms are usually vector control and direct torque control.Although it has good control effect,it is difficult to meet the high performance control under different industrial conditions.Therefore,in order to satisfy the high performance control of PMSM in complex conditions,the Model Predictive Control(MPC)algorithm is gradually applied to the permanent magnet synchronous motor field.However,the MPC algorithm still has its shortcomings,i.e.,the algorithm depends heavily on the motor parameters.When the model parameters are not consistent with the actual parameters,the system will not be able to maintain its good control performance.Therefore,it is particularly important to improve the robustness of MPC parameters.Aiming at the problem of parameter robustness of MPC,3 robust MPC methods with strong parameters are proposed in this paper.The first method is based on the conventional Model Predictive Current Control(MPCC).First,the principle of conventional MPCC is introduced,and the sensitivity of conventional MPCC is analyzed.The influence of system parameters on prediction value is quantified.Then,by establishing the sliding mode observer,the system disturbance is observed,and the system model is modified and the disturbance is compensated to the system.The second way is to change the prediction model to an incremental model,thereby eliminating the flux linkage parameters in the model.Then the disturbance observer is established,and the disturbance is replaced by the PI to calculate the deviation between the model inductor and the actual inductor,and the inductor in the model is corrected gradually,so that the inductor in the model is consistent with the actual inductance.The third method is to build deadbeat inductor observer.On the basis of incremental model,the method uses deadbeat control idea to directly observe system inductance through system disturbance.The inductor is updated in real time to ensure parameter matching and enhance the robustness of parameters.In this paper,three methods are simulated and experimentation.The simulation and experiment verify that the proposed method can reduce the parameter sensitivity of MPC,and enhance the ability of MPC algorithm to resist parameter disturbance when applied to PMSM speed regulation system..