Adaptive controllers for avoidance or evasion in an uncertain environment: Some examples (Q913679)

The problem of obtaining controllers which, for a given class of uncertain dynamical systems, assure that the state of the system avoids a prescribed region of the state space is considered. The uncertain elements may be due to the imperfectly known actions of other agents or an uncertain environment, or both. Each of the controllers considered is a combination of a memoryless controller and an adaptive controller. By considering adaptive controllers the avoidance for a large class of systems is assured. Sufficient conditions are given for the existence of these avoidance controllers. An explicit characterization of the adaptive controllers in terms of appropriate Lyapunov-type functions is also presented. These conditions are utilized to obtain avoidance controllers for a general class of uncertain linear systems and to obtain evasion strategies for planar pursuit-evasion problems in which the only knowledge available on the pursuer's speed is that it is bounded, but not even the bound is known. The pursuer's strategy is one of pure pursuit. These applications are illustrated with numerical simulations.