The maximum principle in optimal control of stochastic differential equations

Project: Research

Project Details

Description

We study relaxed optimal stochastic control problem for systems governed by forward and backward stochastic differential equations of both It and McKean-Vlasov types. The set of admissible controls is a set of measure-valued processes called relaxed controls. We assume that the control variable is allowed to enter both the drift and diffusion coefficients. We prove that these dynamical systems are in fact driven by an orthogonal continuous martingale measure, instead of a Brownian motion. Existence of optimal controls is investigated by using weak convergence techniques and Skorokhod theorem. Necessary conditions for optimal control for these systems in the form of maximum principle are established by means of spike variation techniques. Our results will generalize the Peng's maximum principle to the class of measure valued controls. The extension to stochastic differential equations driven by relaxed Poisson random measures will also investigated.
StatusFinished
Effective start/end date1/04/201/10/22

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.