Finite-Control-Set Model Predictive Current Control For Permanent Magnet Synchronous Motor

Permanent magnet synchronous motor(PMSM)is widely used in various servo control occasions due to its high power density and fast response.For the traditional servo control algorithm,it is difficult to balance various system constraints,and the parameter tuning is complicated.The finite-control-set model predictive control(FCSMPC)algorithm is applied to the current control for PMSM.The proposed algorithms can reduce the computational burden and improve the robustness of the system against parameter mismatch compared to traditional model predictive control.Firstly,the mathematical model of FCS-MPC and the computational solution process of the whole control algorithm are analyzed.The mathematical model of PMSM and inverter is analyzed.On this basis,the FCS-MPC and the current control of PMSM are combined.The prediction model of system is obtained according to the state equation of the motor.The cost function is designed according to the requirements of current tracking and current limiting.The current limiting is integrated into the cost function by a nonlinear function,the Lagrange extrapolation is designed to obtain the future reference value,the predictive controller is used to replace the current loop in traditional field oriented control,so that the current is well tracked and limited.Secondly,because the predictive control is executed in the actual hardware,the calculation burden is large,which leads to the delay of the algorithm application.Therefore,the cause of the delay is analyzed,and the system performance under ideal conditions and delay is analyzed.In order to reduce the influence of delay and improve the steady-state performance,a multi-step predictive control algorithm is introduced.The length of prediction horizon is determined under the condition of comprehensive calculation and steady state improvement.The computational complexity of the standard two-step predictive control is large,and the algorithm is difficult.To reduce the calculation burden,two two-step prediction simplification algorithms based on the deadbeat control algorithm are proposed.The position of reference vector is determined by the deadbeat control to reduce the candidate vector control set,and then applied to the second step prediction to reduce the calculation.The proposed algorithms are compared with the traditional simplified control algorithm to achieve global optimization.Then,the parameter mismatch research of the finite-control-set model predictive control algorithm is carried out.The predictive value of the model predictive control depends on the mathematical model,so the factors affecting the current prediction error are analyzed from the mathematical model,and their effects are qualitatively analyzed from the two perspectives of steady state and dynamics.In addition,for the change of inductance and resistance,The effects of resistance and inductance changes on steady state and dynamic performance are analyzed quantitatively.Two current prediction error compensation schemes are proposed to improve the parameter adaptability of model predictive control and enhance the robustness of the system against parameter mismatch.Finally,the current control with model predictive control algorithm,the multi-step predictive control simplification algorithm and two compensation algorithm based the prediction error are verified by experiments.The experiment shows that the proposed multi-step predictive control can effectively reduce the calculation burden and improve the system performance.The compensation algorithm can also effectively enhance the robustness of the algorithm.