Min-problem for Markov moments and the speed of response (Q806033)

The paper discusses the following problem: given are a constant \(L>0\), smooth functions \(g_ 1,...,g_ k\) and real numbers \(\alpha_ 1,...,\alpha_ n\); find the minimal value \(\tau\) and a measurable function f such that \[ \int^{\tau}_{0}g_ k(t)f(t)dt=\alpha_ k,\quad k=1,...,n;\quad | f(t)| \leq L,\quad t\in [0,\tau]. \] It is shown that in the case of analytic functions \(g_ k\) and small \(\alpha_ k\) it is possible to approximate the solution of the above mentioned problem by means of solutions of similar problems with polynomials instead of the functions \(g_ k\). These results are applied to the minimum time control problem for linear ordinary differential equations.