Chaos detection in chaotic systems with large number of components in spectral domain

A novel technique for chaos detection is proposed by using a time-frequency (TF) representation. The proposed method classifies signals from chaotic oscillators by using entire spectral domain, and it is able to produce accurate results for signals with large number of spectral components in periodic regime. The detector accuracy has been proven on the Colpitts oscillator and logistic map system. The algorithm is tested in the case of moderately noisy signal.