Time series evaluation methods of system dynamics

Abstract

Several system dynamics evaluation methods based on first Lyapunov exponent estimation were developed. Method which eliminates influential periodical component from time-series was presented. Influence of developed methods parameters values on first Lyapunov exponent estimation precision was investigated. Simulation of different linear, nonlinear, stationary and non-stationary systems with the emphasis on the first Lyapunov exponent evaluation was carried out. Sophisticated model of hydrostatic drive control system was analyzed.