ESTIMATION OF CORRELATION DIMENSIONS OF NONLINEAR SYSTEMS UNDER BOUNDED-NOISE EXCITATION

Noise-contaminated dynamic analysis is a complex and difficult topic in nonlinear systems under stochastic background. The random behavior of such systems’ responses may come from measure error, nonlinearity or dynamic noise, etc. The effects of bounded-noise excitation on the Duffing oscillator of Holmes type are discussed. Using the Monte-Carlo method and phase space reconstruction skill, the simulation results of system’s responses and their corresponding correlation dimensions are presented when the parameters in the system assume two different sets of values. It is shown that the presence of external bounded-noise excitation leads to an increase in the correlation dimension of different attractors.