Research On Several Technologies Of Model-free Predictive Control For PMSM Drive System

Low-carbon economy has become the main direction of future development.Under this background,new energy vehicles emerge at the historic moment.New energy vehicles have the advantages of energy saving,emission reduction and environmental protection,which also represent the trend of the world automobile industry.Permanent magnet synchronous motor(PMSM)drive system is the core component and main power source of new energy vehicles,whose control quality directly affects the performance of the electric drive system.Therefore,this paper focus on the problem of inverter nonlinearity,steady-state performance improvement and switching loss reduction of PMSM drive system.In order to enhance the current control performance,the research of model-free deadbeat current control,double-objective finite control set model-free predictive control with discrete space vector modulation(DSVM),hybrid pulse-width modulation(HPWM)based on finite control set predictive control is presented,which has theoretical consequences and practical value for the control performance improvement of PMSM drive system in EVs.Firstly,the model-free deadbeat current control is introduced to solve the problem of parameter uncertainties and inverter nonlinearity in PMSM drive system.By analyzing the dead-time error voltages of dq-axis,the limitation of ultra-local model of model-free current control is pointed out,and the dead-time compensation scheme matched for model-free current control is proposed.The validity of the model-free deadbeat current control with dead-time compensation for PMSM drive system is verified by experimental research.Due to the unsatisfactory steady-state performance of finite control set predictive current control,DSVM is introduced to obtain virtual voltage vectors.Based on the principle of minimum distance of reference voltage vector,the voltage vector space is subdivided into several subsections.The optimal voltage vector can be determined directly according to the subsection of reference voltage vector,which significantly simplifies the optimization algorithm.To reduce the inverter switching frequency,a twolevel decisional cost function is designed to automatically optimize the voltage vector sequence.Based on the above work,double-objective finite control set model-free predictive control with DSVM is proposed.The system experiment proves that the proposed control scheme can achieve the double-objective optimization of current steadystate accuracy and switching frequency reduction of PMSM drive system.HPWM is introduced in the last chapter,and a new finite control set is composed of various PWM strategies.The cost function contains current quality evaluation,switching loss and penalty term of dead-time effect.The optimal PWM strategy is obtained online through the evaluation of cost function.This paper proposes HPWM based on finite control set predictive control,then,the optimization of steady-state current performance and switching loss during the control period could be realized.The feasibility is verified by system simulation.