On GA-based optimal fuzzy control

Two architectures for designing optimal fuzzy control systems are proposed. In both cases, the membership functions in the fuzzy rule bases are tuned by the genetic algorithms. The objective is to explore a fuzzy controller by minimizing a quadratic cost function. In the first architecture, the employed controller is a conventional fuzzy logic controller which uses the system states as input variables. Consequently, the reciprocal of the cost function to be minimized could be directly applied towards evaluating the fitness of the controller. In the second architecture, a fuzzy sliding mode controller is adopted. The combined information of the system states, i.e. the sliding function, form a single input variable. The problem of minimizing the cost function in this case could be transformed to that of deriving an optimal sliding surface. Then, a faster hitting time and a smaller distance away from the sliding surface a controller has, a higher fitness it gets. Simulations and comparisons are taken on both cases