Maximum Principle for General Controlled Systems Driven by Fractional Brownian Motions

TL;DR

This paper establishes a maximum principle for optimal control of systems driven by fractional Brownian motions with Hurst parameter greater than 1/2, using Malliavin calculus and stochastic analysis.

Contribution

It develops a new maximum principle involving backward stochastic differential equations and Malliavin derivatives for systems driven by fractional Brownian motions.

Findings

Derived a maximum principle for fractional Brownian motion-driven systems.

Extended classical maximum principle to systems with partial information.

Developed new stochastic analysis tools for fractional systems.

Abstract

We obtain a maximum principle for stochastic control problem of general controlled stochastic differential systems driven by fractional Brownian motions (of Hurst parameter $H>1/2$). This maximum principle specifies a system of equations that the optimal control must satisfy (necessary condition for the optimal control). This system of equations consists of a backward stochastic differential equation driven by both fractional Brownian motion and the corresponding underlying standard Brownian motion. In addition to this backward equation, the maximum principle also involves the Malliavin derivatives. Our approach is to use conditioning and Malliavin calculus. To arrive at our maximum principle we need to develop some new results of stochastic analysis of the controlled systems driven by fractional Brownian motions via fractional calculus. Our approach of conditioning and Malliavin calculus is also applied to classical system driven by standard Brownian motion while the controller has only partial information. As a straightforward consequence, the classical maximum principle is also deduced in this more natural and simpler way.