Application of iterative learning control to active power filter and its robustness design

This paper presents the application of iterative learning control scheme to active power filter (APF) with hysteresis current shaping technique. The iterative learning control strategy for APF is incorporated with combination of feed forward and feedback learning control. The feed forward learning control is designed for accomplishment of recognition of steady state model. And the feedback control, acting as an iterative learning filter, is focused on the compensation for its dynamics and phase characteristics. Further, to meet the requirement of robustness and tracking speed, the strategy to self-optimize the control system, which is realized by adjusting the learning control law and improving dynamic characteristics during learning cycle, is developed. Based on the repetitive feature of the system, control to reduce THD of source current is established through adjustment of iterative learning control laws. All these findings are supported by some simulation and experiment results.