Partially-Observed Maximum Principle for Backward Stochastic Differential Delay Equations
Abstract
Dear Editor, This letter investigates a partially-observed optimal control problem for backward stochastic differential delay equations (BSDDEs). By utilizing Girsanov's theory and convex variational method, we obtain a maximum principle on the assumption that the state equation contains time delay and the control domain is convex. The adjoint processes can be represented as the solutions of certain time-advanced stochastic differential equations in finite-dimensional spaces.