Markov Parameters Evaluation By Integral Methods

Abstract

This paper deals with a method for Markov parameters evaluation by integral methods for proper dynamical systems. The step response of the system is processed and a set of weighted integrals, performed on finite time intervals, allows creating an algebraic system of equations whose unknowns are the Markov parameters. Theoretically this is an infinite dimensional system but it is truncated to a finite order one. The convergence conditions and truncation error effects are analyzed. Because the Markov parameters are coefficients of Laurent series it is proposed a techniques for series convergence domains evaluation.