基于扇区判断的三电平APF多目标模型预测电流控制

Current Control Prediction of Three-level APF Multiple Target Model Based on Partition Judgment

有源滤波器是改善电网电能质量的关键设备之一。针对三电平有源滤波器传统模型电流预测控制方法中滚动优化次数多,程序运行时间长,开关损耗高的问题,研究了一种基于扇区判断的三电平有源滤波器多目标模型预测电流控制方法,首先对参考电压进行扇区的判断,然后再确定所在扇区的最优开关状态子集,以此进行滚动优化,同时考虑开关损耗的问题,在价值函数中引入平均开关频率约束。该方法不仅可以减少程序运行时间,同时在不影响补偿效果的情况下降低了平均开关频率。最后通过仿真和实验验证了所研究方法的正确性和有效性。

Active power filter is one of the key devices to improve power quality of power network. As for such problem as multiple number of rolling optimization,long operation time of program and high switching loss in the traditional model current prediction and control method of three-level active fiher,a kind of current prediction and control method of three-level active filter multiple target model based on partition judgment is studied. Firstly, partition judgment for the reference voltage is performed, then the sector of the optimal switching state subset is determined and,on this basis,to perform rolling optim- ization and, at the same time ,with consideration of switch loss ,the average switching frequency constraint is introduced in the value function. This method can not only reduce operation time of the program,but al- so reduce the average switching frequency at the condition of not affecting the comp-ensation effect. Finally ,the correctness and effectiveness of the method are verified by simulation and experiment.