A new genetic based approach to fuzzy controller design and its application

One of the major challenges in the current fuzzy control research is the automatic design of multiple input controllers for complex nonlinear systems. This paper presents a new genetic-based scheme to treat this issue: the so-called premise learning approach. We propose to search in the input domain for suitable rule premises. The rule premises are coded in a general way allowing AND- as well as OR-connections of the linguistic terms, in combination with a certain class of input and output fuzzy sets. The rule structure and the fuzzy sets are optimized by the genetic algorithm at the same time. With this new approach a considerable reduction of the number of necessary rules may be expected. This method is used to design a fuzzy controller to balance an inverted pendulum. Simulations as well as results of a real laboratory plant are shown to demonstrate the effectiveness of the new method.