Chaos in human rhythmic movement

Rhythmic movements typical of locomotory actions are usually modeled as limit cycle dynamics, and their deviations from pure periodicity are attributed to stochastic physiological noise. In the present study, the dynamics of human rhythmic movements were found to contain more than the 2 dynamically active variables expected from limit cycle dynamics; the number depended upon the size of the limb oscillator. Observed positive Lyapunov exponents and fractal attractor dimensions indicated that the gross variability of human rhythmic movements may stem largely from low-dimensional chaotic motion on strange attractors.