Using learning samples to construct fuzzy logic systems with the application to inverted pendulum control

Fuzzy logic systems (FLS) are nonlinear systems and have a very strong function approximation ability. As to achieve FLS´s function approximation ability, it is very important to find algorithms to synthesis the fuzzy rule parameter value. The table-lookup algorithm is a kind of fuzzy rule design method and its computational cost is quite low. Based on the classic table-lookup algorithm idea, we designed another simpler table-lookup synthesis method, called a simplified table-lookup algorithm, and its computational cost is less than the original one. Then we used this algorithm to design the rule parameters of a standard fuzzy controller, which has four inputs and one output. Finally, we applied this controller to inverted pendulum control applications. The results show that our simplified algorithm can realize a stabilization controller.