Research On Finite Set Model Predictive Control Of Open-winding Permanent Magnet Synchronous Motor

The open winding permanent magnet synchronous motor has become the focus of research because of its advantages in multi-level operation characteristics and faulttolerant control.Open-winding motors fed by dual inverters have two connection methods: common bus and independent bus.Both have redundant switching states.The difference is that there is zero-sequence circulating current in the common bus topology,which must be suppressed.Meanwhile,model predictive control has a strong application potential to solve the unique multi constraint problem of dual inverter system.Therefore,a model predictive control strategy has been developed in this paper for open-winding permanent magnet synchronous motors.The mathematical model of the OW-PMSM system in different coordinate systems was first established,and the topology of the dual inverter under two wiring methods was studied.Then,based on its structural characteristics and circuit model,a simulation model of the OW-PMSM system was built in SIMULINK and tested on this basis.It lays the foundation for the following research on the FCS-MPC algorithm applicable to openwinding motors.To apply the traditional FCS-MPC strategy to the dual power OW-PMSM system,the mechanism of prediction current error between model parameters and actual model,and the influence of modeling error on FCS-MPC algorithm are analyzed.Based on these,the FCS-MPC strategy suitable for OW-PMSM system is proposed.Besides,the lengthy prediction process is one of the common problems of model prediction algorithms.For this,a simplified model prediction voltage control strategy is proposed in this paper.This strategy combines the idea of deadbeat which contains two-sector judgments on the reference voltage vector,and only two evaluations and one comparison are needed to get the optimal voltage vector,significantly improving the efficiency of optimization.Furthermore,the problems of one-to-many of vector and switch states of the dual power OW-PMSM system have also been solved by the function of the number of times of switching.For the problem of zero-sequence current the single-supply OW-PMSM system,an FCS-MPC strategy based on zero-sequence current suppression is proposed.This strategy utilizes the characteristics of model predictive control that can handle multiple control targets and system constraints.By adding a zero-sequence voltage control term to the value function,the purpose of suppressing zero-sequence current is achieved.To improve the steady-state performance of the system and solve the problem that the switching frequency of the system is not fixed,a finite switching sequence model predictive control strategy for OW-PMSM system is proposed.On the premise of eliminating the switch state with zero-sequence voltage and following the principle of minimum switch frequency,the finite set is determined as six groups of switch sequences.The action time of each group of switch sequences is calculated in turn by deadbeat control method,and the optimal switch sequence is obtained by online optimization finally.