On a problem of calculating the solvability set for a linear system with uncertainty (Q6147958)

The authors consider a linear-convex control system defined by \[ \dot{x}(t)=A(t)x(t)+B(t)u+C(t)v, t \in [t_{0},t_{1}] , \] where \(A(t), B(t)\) and \(C(t)\) are continuous matrix. The controls \(u(t)\) and \(v(t)\) belong, respectively, to the sets \(P(t)\) and \(Q(t)\). The set of control functions \(P(t)\) and \(Q(t)\) are continuous set-valued mappings taking values in the set of convex compact sets (continuity is understood in the sense of Hausdorff metric). The system may have uncertainties the possible values of which are subject to strict pointwise constraints. For this system, over a finite period of time, taking into account the constraints, the authors study the problem of guaranteed hitting the target set from a given initial position despite the effect of uncertainty. The primary stage in solving the problem involves the construction of an alternating integral and a solvability set. In fact, the main results are presented in the two-dimensional context for which a method is proposed for finding the solvability set without the need to calculate the convex hull of the difference of the support functions of the sets.