基于离散空间矢量调制的永磁同步电机无参数预测电流控制OA北大核心CSTPCD

Traditional parameter-free predictive current control(PFPCC)based on current gradient update has these problems of stagnant current gradient update and large current ripple.To solve these problems,this paper proposes an optimized PFPCC method based on discrete space vector modulation(DSVM).Firstly,by analyzing the current gradient relationship of different voltage vectors on the α-β axis,the mathematical relationship between each voltage vector and current gradient in two adjacent control periods is obtained.Then,the current gradients of all voltage vectors are updated in one control cycle,which effectively reduces the stagnation effect in traditional PFPCC.In order to further reduce the current ripple,DSVM is introduced into the proposed method.Here,combined with the vector selection method of DSVM,all current gradients can be updated with a small amount of calculation,thus ensuring the reliability and accuracy of current prediction.The experimental results show that the proposed PFPCC method has similar dynamic and static performance compared with model-based predictive current control(MBPCC).Compared with single-vector PFPCC,the DSVM-PFPCC can obviously reduce the current ripple,and improve the control performance in the actual system while ensuring the dynamic and static performance.