Abstract:
Several problems of dynamic systems control can be reduced to geometric games. The problem of stabilization is an example. In this paper the criteria of a saddle point in a geometric game is proved under more general conditions than earlier. Algorithms for finding of a saddle point are given in cases where the strategy set of one of the players is (1) a ball in $\mathbb R^n$, (2) a closed interval, (3) a polyhedral, and the strategy set of the other player is an arbitrary convex set.
Fulltext:
PDF file (224 kB)