Stochastic Periodic Solutions for Newtonian Systems via Lyapunov Function

TL;DR

This paper proves the existence of stochastic periodic solutions for Newtonian systems with time-periodic forcing using Lyapunov functions, confirming a longstanding conjecture and addressing open problems in the field.

Contribution

It introduces a new existence theory for stochastic periodic solutions in Newtonian systems, confirming the stochastic Levinson conjecture and solving open problems related to dissipative systems.

Findings

Existence of solutions under polynomial growth conditions.

Persistence of solutions under bounded perturbations.

Confirmation of the stochastic Levinson conjecture.

Abstract

This paper establishes an existence theory for distributed periodic solutions to Newton's equation with stochastic time-periodic forcing, where the friction matrix is the Hessian of a twice continuously differentiable friction function. Employing the Khasminskii criterion and Lyapunov functions, we prove its existence under the assumptions that both the friction and potential functions tend to positive infinity at infinity, and their gradient inner product grows at least like an even-power polynomial. Provided the potential grows sufficiently fast at infinity, this result persists under bounded friction matrix perturbations. This largely confirms the stochastic Levinson conjecture proposed in [8] and resolves the open problem and the interesting problem posed in [8, P. 342]. Distributed periodic solutions exist in two key cases: 1. Polynomial Case: Both functions are polynomials with positive definite leading terms (a minimal condition for dissipation). 2. Plasma Physics Case: The potential force is bounded, while the friction function has sufficiently high growth (as in many plasma physics models).