There has been an increased attention on the intelligent control theory, which is motivated by human/biological control behavior in recent years. The intelligent control has been revealed to be appropriate to an uncertain dynamic system through some case studies and computer simulations. However, the stability of the intelligent control system is not clear due to the inherent nonlinearity of the intelligent control, e.g., the fuzzy logic control and the neural network control theory. It may be the main factor that one hesitates to adopt the intelligent control in critical environments.
In this thesis, the robustness/stability of fuzzy logic control is investigated by the well-known Lyapunov system theory. It is based on the observation that the pattern of control input is similar between a conventional robust controller and a prevalent fuzzy logic controller. Furthermore, by compensating system uncertainty using an intelligent model, the control efficiency is shown to be greatly increased. Lastly, a complete intelligent control algorithm is proposed by combining the fuzzy logic controller and the intelligent model for uncertain dynamic system. The stability of the closed loop control system is proved in the sense of Lyapunov. These algorithms are verified by computer simulations for a 2-link robot manipulator.