The posture stability driving control of a human riding-type unicycle has been realized. The robotic unicycle is considered as a biomechanical system using an internal world representation with a description of emotion, instinct and intuition mechanisms. We introduced intelligent control methods based on soft computing, and confirmed that such an intelligent control and biological instinct as well as intuition together with a fuzzy inference is very important for emulating human behaviors or actions. For the fitness function of the generic algorithm, a new physical measure of the minimum entropy production for a description of the intelligent behavior in a biological model is introduced. The calculation of robustness and controllability of the robotic unicycle is presented. This paper provides a general measure to estimate the mechanical controllability both qualitatively and quantitatively, even if any control scheme is applied. The measure can be computed using a Lyapunov function coupled with the thermodynamic entropy change. Fuzzy simulation and experimental results of a robust intelligent control motion for the robotic unicycle are discussed.
展开▼