Fokker-Planck Equation and Feynman-Kac Formula for Multidimensional Stochastic Dynamical Systems with Lévy Noises and Time-Dependent Coefficients

Qingyan Meng,Yejuan Wang,Peter E. Kloeden,Xiaoying Han
DOI: https://doi.org/10.1016/j.matcom.2024.10.014
IF: 3.601
2025-01-01
Mathematics and Computers in Simulation
Abstract:The aim of this paper is to establish a version of the Feynman-Kac formula for the time- dependent multidimensional nonlocal Fokker-Planck equation corresponding to a class of time-dependent stochastic differential equations driven by multiplicative symmetric (or asymmetric) a-stable L & eacute;vy noise. First the forward nonlocal Fokker-Planck equation is derived by the adjoint operator method, overcoming the challenges posed by time-dependent multidimensional nonlinear symmetric a-stable L & eacute;vy noise. Subsequently, the Feynman-Kac formula for the forward multidimensional time-dependent nonlocal Fokker-Planck equation is established by applying techniques for the backward nonlocal Fokker-Planck equations, which is associated with the backward stochastic differential equation driven by the multiplicative symmetric a- stable L & eacute;vy noise. Notably, in the case of asymmetric a-stable L & eacute;vy noise case, the presence of the characteristic function in the nonlocal operator adds complexity to the analysis. Using the Feynman-Kac formula, it is demonstrated that the solution of the forward nonlocal Fokker- Planck equation can be readily simulated through Monte Carlo approximation, especially in scenarios involving long-time simulation settings with large steps. These concepts are illustrated with intriguing examples, and the dynamic evolution of the probability density function corresponding to the stochastic SIS model and the stochastic model for the MeKS network (reflecting the interactions among the MecA complex, ComK and ComS) are investigated over an extended period.
What problem does this paper attempt to address?