Lyapunov theory based intelligent fuzzy controller for Inverted Pendulum

Proposed work is aimed at designing an intelligent fuzzy controller with guaranteed stability. We employ Lyapunov theory on the linguistic fuzzy rules to constrain action set for the fuzzy controller. Lyapunov theory constrained control action results in an intelligent controller that exhibits stability even under disturbances. Proposed controller is suitable for continuous state action spaces as it employs fuzzy systems for generalization in a reinforcement learning (RL) framework. We employ proposed controller on the benchmark task of balancing of an Inverted Pendulum under disturbances. Simulation results elucidate practicality and reliability of the proposed control scheme.