Intelligent fuzzy controller design on WARP for mechanical systems with complex dynamic behaviors

The methodology of intelligent fuzzy controller design is introduced. The basis of the methodology is simulation of computer dynamics for mechanical systems with the help of qualitative physics and the search of possible solutions with genetic algorithm (GA). On artificial neural networks optimal solutions are defined and knowledge base of fuzzy controller on the Weight Associative Rule Processor (WARP) is formed. Applications of this methodology on examples including a mobile robot for service use, a wall climbing robot (WCR) for decontamination of nuclear-power plant, a DC/DC converter and an apparatus for artificial ventilation of lungs, are examined.<
>