Chaos and Arnold Diffusion in Dynamical Systems

Salam, F. M. A., J. E. Marsden and P. P. Varaiya

IEEE Trans. Circuits and Systems, 30, (1983), 697-708

Abstract:

Chaotic motion refers to complicated trajectories in dynamical systems. It occurs even in deterministic systems governed by simple differential equations and its presence has been experimentally verified for many systems in several disciplines. A technique due to Melnikov provides an analytical tool for measuring chaos caused by horseshoes in certain systems. The phenomenon of Arnold diffusion is another type of complicated behavior. Since 1964, it has been playing an important role for Hamiltonian systems in physics. We present a tutorial treatment of this work and its place in dynamical systems theory, with an emphasis on results that can be checked in specific systems.
A generalization of the Melnikov technique has been recently developed to treat n -degree of freedom Hamiltonian systems when n > 3 . We extend the Melnikov technique to certain non-Hamiltonian systems of ordinary differential equations. The extension is made with a view to applications in the physical sciences and engineering.