McKean-Vlasov forward-backward doubly stochastic differential equations and applications to stochastic control

TL;DR

This paper establishes the existence and uniqueness of solutions for McKean-Vlasov forward-backward doubly stochastic differential equations in infinite-dimensional spaces and applies these results to stochastic control problems.

Contribution

It provides the first analysis of MV-FBDSDEs in infinite-dimensional spaces and connects these equations to stochastic control via the maximum principle.

Findings

Proved existence and uniqueness of solutions for MV-FBDSDEs

Extended results to stochastic maximum principle applications

Provided examples and counterexamples to illustrate the theory

Abstract

This paper investigates first the existence and uniqueness of solutions for McKean-Vlasov forward-backward doubly stochastic differential equations (MV-FBDSDEs) in infinite-dimensional real separable Hilbert spaces. These equations combine the features of forward-backward doubly stochastic differential equations with the mean-field approach, allowing the coefficients to depend on the solution distribution. We establish the existence and uniqueness of solutions for MV-FBDSDEs using the method of continuation and provide an example and a counterexample to illustrate our findings. Moreover, we extend the practical applicability of our results by employing them within the context of the stochastic maximum principle for a control problem governed by MV-FBDSDEs. This study contributes to the field of stochastic control problems and presents the first analysis of MV-FBDSDEs in infinite-dimensional spaces.