Self-tuning controller with fuzzy filtered-X algorithm

A fuzzy filtered-X algorithm is proposed which has sell-tuning organization under active noise control. By imitating human learning processes, the fuzzy control algorithm does not require a plant model to make its control rules. Instead, rules are generated by using the history of the input-output noise pairs. The new approach not only provides better performance than the conventional filtered-X algorithm but also leads to a reduction in the complexity of the active noise control system.