Optimal design of a multi-level fuzzy controller using a multi-objective particle swarm optimization algorithm

In the present study, a novel multi-level fuzzy control scheme is introduced and optimized by a multi-objective algorithm to determine Pareto fronts of conflicting objective functions. Against traditional techniques, a fuzzy controller is designed to stabilize the system, while a supervisory controller is employed to handle the uncertainties. In order to challenge the performance of the suggested idea, it is applied for stabilization of a nonlinear under-actuated two degree-of-freedom inverted pendulum system. The minimized objective functions are the normalized error of the pendulum angle as well as the normalized error of the cart position. Results on the found non-dominated solutions and system states clearly illustrate the superiority of the proposed optimization algorithm and designed control method in comparison with the well-known approaches introduced in literature.