The asymptotical behavior of the solution to the time-optimal problem for a linear system under perturbation of initial data (Q1385859)

The paper deals with the time-optimal problem of a controlled object whose dynamics is given by the following differential equation: \[ \dot{x}=Ax+Bu, \qquad x \in \mathbb{R}^{n},\quad u \in \mathbb{R}^{m} \tag{*} \] with constrained control vector function \(| | u| | \leq 1\), where \(| | \cdot| | \) is the Euclidean norm. The time-optimal problem is to transfer the starting point \(x(t_{0})=x_{0}+\varepsilon y\) to the target, taken as the origin, i.e. \(x(\theta) = 0\), in minimum time \((\theta - t_{0}) \rightarrow \min\). There has been assumed that if \(\varepsilon = 0\) then the system (*) has a solution induced by a discontinuous control function. In the other case, the control function is continuous. The paper shows that it is possible to expand each solution into some asymptotic series which ensures an accuracy to any power of the parameter \(\varepsilon\).