The Existence of Shilnikov Homoclinic Orbits in the Michelson System: A Computer Assisted Proof

Abstract

In this paper we present a new topological tool which allows us to prove the existence of Shilnikov homoclinic or heteroclinic solutions. We present an application of this method to the Michelson system y + y + 0.5y 2 = c 2 [16]. We prove that there exists a countable set of parameter values c for which a pair of the Shilnikov homoclinic orbits to the equilibrium points (±c √ 2, 0, 0) appear. This result was conjectured by Michelson [16]. We also show that there exists a countable set of parameter values for which there exists a heteroclinic orbit connecting the equilibrium (−c √ 2, 0, 0) possessing a one-dimensional unstable manifold with the equilibrium (c √ 2, 0, 0) possessing a one-dimensional stable manifold. The method used in the proof can be applied to other reversible systems.To verify the assumptions of the main topological theorem for the Michelson system, we use rigorous computations based on interval arithmetic.

Stationary Coexistence of Hexagons and Rolls via Rigorous Computations

“…Second, a distinct type of phasespace oriented, topological computer-assisted approach (see [19,20,21,22] and the references therein), can be applied to prove the existence of the connections. Such methods have been used successfully to find connecting orbits in a variety of nonlinear systems [23,24,25,26,27], and they are especially adept in low regularity settings. In the present paper, on the other hand, by combining a functional analytic setting and a parameterization method, we exploit the high regularity of solutions in the (parabolic) pattern formation problem.…”

Covering relations, cone conditions and the stable manifold theorem

“…⇒ N 0 of increasing length using the compactness argument one can show (see [34,Col. 3.10]) that for every y ∈ B s there exists x ∈ B u , such that (68) holds for z = c −1 N 0 (x, y).…”

Analytic Continuation of Local (Un)Stable Manifolds with Rigorous Computer Assisted Error Bounds

“…The CAPD algorithms are based on the pioneering work of Lohner [22,23,24], and instead of using fixed point arguments in function space to manage truncation errors, develop validated numerical bounds based on the Taylor remainder theorem. The CAPD algorithms provide results in the C k category, and are often used in conjunction with topological arguments in a Poincare section [25,26,27,28,29,30] to give computer assisted proofs in dynamical systems theory.…”