Fuzzy Finite Set Model Predictive Current Control For PMSM Based On Improved DSVM

Model Predictive Control(MPC)is used in Permanent Magnet Synchronous Motor(PMSM)due to its simplicity,multi-objective optimisation and fast dynamic response.Control strategy plays a key role in areas with high performance requirements such as oil drilling,aerospace industry and electric vehicles.However,the traditional Finite Control Set-Model Predictive Control(FCS-MPC)has a large error between the reference current and the predicted current due to the inherent discrete nature of the power inverter and the fact that only one voltage vector can be selected in a single switching cycle,resulting in a large error between the reference current and the predicted current and a large current ripple due to the harmonic current component.To address this problem,a PMSM fuzzy finite set model predictive current control strategy based on improved Discrete Space Vector Modulation(DSVM)is proposed in this paper.The improved control strategy combines a DSVM technique with fuzzy dynamic cost function weighting coefficients to improve the current loop controller of the traditional finite set model predictive control strategy to reduce the current ripple.The FCS-MPC strategy introduces the DSVM technique,where multiple fundamental voltage vectors are applied in a single switching cycle,and a virtual voltage vector is linearly synthesised from multiple fundamental voltage vectors to provide more optional voltage vectors,Therefore,a Discrete Space Vector Modulation-Model Predictive Control(DSVM-MPC)based on discrete space vector modulation is proposed,which effectively reduces the error between the reference and predicted currents and achieves a reduction in current ripple.To avoid the exhaustion of a finite number of available voltage vectors,the improved solution in this paper uses a two-step finding method to select the optimal voltage vector in two steps.The two-step finding method reduces the number of voltage vectors in the prediction process from 38 to 14,and the selected voltage vector is directly found to the corresponding duty cycle value to generate the inverter pulse signal for input to the inverter.In addition,to further reduce the current ripple,the configuration relationship between the weights of the variables of the motor control system in different states and the weight coefficients of the current term of the cost function is analysed,and the configuration relationship is designed as a relevant scheme such as a fuzzy inference rule,which enables the dynamic optimisation of the weight coefficients of the cost function through a fuzzy controller to further reduce the current ripple.To compare and verify the experimental findings of the control strategies,simulation models of the FCS-MPC strategy,DSVM-MPC strategy and the improved control strategy were developed and based on this,the simulation experiments were verified by the Matlab/Simulink simulation system and the hardware experiments by the Links-RT semi-physical simulation system.The simulation and hardware experimental results show that the improved control strategy reduces the current ripple in steady state and reduces the computational effort increased by the DSVM technique.The improved control strategy has similar dynamic characteristics to the FCS-MPC and DSVM-MPC strategies,while suppressing the current ripple.A comprehensive analysis of the experiments with the three control strategies shows that the improved control strategy reduces current ripple compared to the FCS-MPC strategy,its computational burden is reduced compared to the DSVM-MPC strategy and its control performance is improved compared to the other two control strategies.