A soft computing-based mathematical model of an intelligent control system for a complex, globally unstable object

In this paper, the concept of an intelligent control system for the complex nonlinear biomechanical system of a robotic unicycle is presented. The results of stochastic modeling of the fuzzy intelligent control system are demonstrated under various types of external and internal random perturbations for a dynamic, globally unstable object represented by a robotic unicycle. The thermodynamic approach to the investigation of optimal control processes in complex nonlinear dynamic systems is employed. A relationship between the entropy production and the Lyapunov function (the measure of stochastic stability) is established. A new approach to designing intelligent control systems, which is based on the principle of minimum entropy production during the motion of the control object and in the control system itself, is developed. The minimum of entropy production is the fitness function of a genetic algorithm, which allows one to construct a robust control for a robotic unicycle.