Fuzzy Iterative Learning Control for Three-Phase Shunt Active Power Filters

In this paper, we study the power-factor correction problem with shunt active power filter (APF) to compensate the harmonics and reactive power due to nonlinear load and switching devices. Since the current reference commands are periodical and drastically vary with time, a current-cycle iterative learning control is adopted to achieve the current tracking objective. Fuzzy rules to tune the learning gains are also devised, which act to increase robustness and improve tracking performance of the iterative learning controller. The proposed approach can provide good tracking accuracy and fast convergence. The experimental results show that with the proposed fuzzy iterative learning controller, reduced THD in source currents and high power factor are established. Fast dynamical response to step change in load is also observed