Improved Model Predictive Control Of Permanent Magnet Synchronous Motor Based On Fuzzy Algorithm

In the permanent magnet synchronous motor(PMSM)systems,speed-current double closed loop is a common control structure.Among them,the performance of the current loop directly determines the current loop dynamic performance and steady-state performance of the motor system.It is the key to improving control performance of a PMSM system.Predictive current control(PCC),which applies the model of motor and inverter to predict the current response in the next control interval,can broaden the bandwidth of current loop and improve its transient response.So it is being widely researched.According to the application mode of voltage vector,predictive current control can be categorized into direct predictive control(DPC)and PWM predictive control(PPC).Current predictive control can make the motor current get good dynamic and steady state response,but there are some problems.Because predictive control is a control method based on the model.The controller has higher requirements on the model accuracy.However,some of the key PMSM system parameters in the actual system are difficult to obtain accurately,and some parameters may change with the running state of the motor.Such as stator inductance,resistance,flux linkage and so on.The above conditions lead to the control model parameters mismatch.Motor model parameters mismatch will make the current control oscillation or static error.Current oscillation will lead to motor mechanical shock and drive over-current damage,and current static error will lead to the torque mode can not output reference torque and other issues.The above parameter sensitivity issues are particularly prominent in PWM predictive control.Aiming at the parameters sensitivity issues of PWM predictive control,an improved model predictive control based on fuzzy control algorithm is proposed in this paper.The intelligent control algorithm is introduced to judge the dynamic,steady state processes and parameters mismatch.Model predictive controller is compensated by proportional integral(PI)link with Weights.In the dynamic process,the PI compensation link effect is reduced,in order to reduce the influence of integral action on the dynamic performance of model predictive control.In the steady-state process the role of PI compensation link is enhanced,so as to eliminate the problems of oscillation and static error brought by the model predictive control parameters mismatch.At the same time,a novel flux observer is designed based on fuzzy algorithm.The introduction of flux observer fundamentally solved the influence of flux error on the control performance.Finally,the experimental verification is carried out on the 2.3kW permanent magnet synchronous motor experimental platform.The improved algorithm and the traditional algorithm are compared and tested.The results show that the proposed algorithm guaranteed the dynamic performance of the model under the condition of rated parameters.And it also eliminated the influence of the flux parameters on the dynamic and steady-state performance and the influence of the inductance parameters on the steady-state performance.The parameter robustness of the system is enhanced.