Problem of adaptive minimax control for pursuit-evasion process

The problem of adaptive minimax control for the pursuit-evasion process is considered in discrete-time dynamical systems which consist of two controlled objects described by linear recurrent vector equations. It is assumed that the sets constraining all a priori undefined parameters axe known, as well as the history of realizations of the control signal of object I controlled by the pursuer, and of a signal that carries incomplete information about object II, controlled by the evader. For organizing a minimax control of pursuit in a chosen class of admissible adaptive control strategies, a recurrent procedure is proposed, each step of which is based on realization of an posteriori minimax nonlinear filtering process and on solving linear and convex programming problems. The results obtained in this report can be used for computer simulation of actual dynamical processes and for designing controlling and navigation systems. Copyright (C) 2001 IFAC.