Chaos Study of the Lamprey Neural System via Improved Small Dataset Method

Abstract

This paper is concerned with the locomotion property of the Lamprey neural system that is modeled by the winnerless competition (WLC) networks. An improved small dataset method for computing the largest Lyapunov exponent is proposed and applied to chaos detection. Application to classical non-linear systems shows that the new algorithm not only works effectively but also achieves better accuracy than the Wolf method. The new algorithm is then employed to study the chaotic properties of the Lamprey neural system. In addition, phase portrait for small perturbation on initial states of the dynamic system is also drawn to aid in chaos determination. Simulation results demonstrate that under some mild external stimulus, the Lamprey neural system exhibits chaos, when external stimulus continues increasing, the Lamprey neural system could return back to steady state.