On a generalization of one game control problem in the class of finitely additive measures (Q619463)

The authors consider a linear dynamic game control problem with piecewise constant and right continuous control actions \(v(t)\) and \(u(t)\). System dynamics is described by a linear time varying scalar state equation with discontinuous control coefficients which is transformed into an integral form. The control objective is to maximize (with respect to \(v\)) the minimal (with respect to \(u\)) value of the scalar function of the terminal state of the system. It is assumed that maximization precedes minimization. Since generally the considered maximin problem may have no accurate solutions the authors construct an extension that ensures the existence of solutions close to optimal ones in the case when the objective function is given with a small error. More precisely, the generalization of the game control problem is based on the use of finitely additive measures satisfying some impulse constraints. The line of reasoning follows the one proposed in previous studies of the first author (see e.g. [\textit{A. G. Chentsov}, Asymptotic attainability. Transl. from an orig. Russian manuscript. Mathematics and its Applications (Dordrecht). 383. Dordrecht: Kluwer Academic Publishers. xiv, 322 p. (1997; Zbl 0859.93002]).