Research On Model Predictive Control Of Permanent Magnet Synchronous Motor

Permanent magnet synchronous motor(PMSM)has been widely used in aerospace,numerical control machine tool and renewable energy automobile.However,the PMSM drive system is a strong coupling and nonlinear system.In order to improve the control performance of PMSM drive system,this dissertation puts forward specific solutions from reducing time complexity,improving steady state performance,enhancing fault tolerant ability and robustness of PMSM driver system,respectively.This dissertation has relevant theoretical contributions and engineering research values.Firstly,basic mathematical model of PMSM drive system based on three-levels neutral point clamped inverter is established.Advantages and disadvantages of conventional finite control set model predictive control method and modulated model predictive control method are analyzed.In order to overcome disadvantages of conventional modulated model predictive control method,which is complicated and is difficult to run in low cost processor,an improved low complexity modulated model predictive control method is proposed.The proposed method adopts predictive voltage equations instead of predictive current equations,which can reduce the time complexity effectively.The equivalence of modulated model predictive control method based on voltage equations and conventional modulated model predictive control method is proved by mathematical method.To improve the steady state performance,the dwell times of the synthesis vector are calculated by using the traditional volt-second balance principle,which reduces total harmonic distortion of currents effectively.Secondly,the problem of inaccurate calculation of dwell times in conventional modulated model predictive control method is analyzed theoretically,and an optimal duty cycle modulated model predictive control method is proposed.The cost function is defined by sum of squares of difference between predictive voltage vector and synthetic voltage vector.The duty cycle can be calculated by solving the cost function.In order to compare with the conventional finite control set model predictive control method and modulated model predictive control method at high sampling frequency,these two predictive methods are improved to enable them to operate at high sample frequency.Thirdly,in order to enhance fault tolerant ability of PMSM drive system.An eight-switch three-phase inverter topology is introduced,and a fault tolerant modulated model predictive control method based on this reconstructed topology is derived,which enhances the fault tolerant ability of PMSM driver system.At the same time,a comparative study is made between the conventional finite control set model predictive control method and the proposed method on eight-switch three-phase inverter topology.Simulation and experimental results validate the effectiveness of the proposed method.Finally,in order to enhance the robustness of PMSM driver system,the disturbances of PMSM driver system are analyzed theoretically,and a modulated model predictive sliding mode composite control method is proposed.Modulated model predictive control method is used for the inner loop and the speed loop is a sliding mode control method which combines improved sliding mode reaching law and an disturbance torque observer.The load torque disturbance observer is used to compensate the influence of external disturbance,and the stability of speed loop is proved by using Lyapunov theory.The contradiction between chattering and reaching velocity in conventional sliding mode control is solved and robustness of PMSM drive system is improved effectively.