Anfis-based Inverted Pendulum Control

Inverted Pendulum system is usually used as the controlled object Stabilization of control is investigated. Control methods of Inverted Pendulum are summarized. The mathematical models of single, double and triple Inverted Pendulum are built. The controllability near the balanced position is given respectively. A new concept of FNN(Fuzzy Neural Network) control is presented. Method based on state variables synthesis is used to decrease the dimensions of the controller. The problem of rule explosion in FNN controller is solved for multi-variable systems. Some parameters in the controller are optimized by ANFIS(Adaptive Neural-Fuzzy Inference System). Method designed is applied in not only simulation for single, double and triple Inverted Pendulum but also real control for single and double systems. The results indicate that the method has better performance than LQR(Linear Quadratic Regulator) and traditional fuzzy control. The experiments of anti-disturbance show its robustness.