Identification Method of the Time-Series Properties of the Aviation Events

Abstract

The new empirical method was developed for estimating the relative contribution of the deterministic, chaotic and noise components in the time series of the accident rate of aircraft in the Armed Forces of the Russian Federation. The behavior analysis of the senior Lyapunov exponent dependencies on the parameter of “number of neighbors used” specified in the algorithm in the case of model series (deterministic, chaotic and colored noise series) enabled to estimate the boundaries of numerical values of the parameters typical for these types of series and then apply them to the noise percentage analysis in the real series of the aviation events number.