pipette
ESEspañol

Vector- and operator-valued backward stochastic equations with finite-variation drivers and a maximum principle for singular stochastic control in infinite dimensions

Ying Hu, Guomin Liu, Shanjian Tang

Preprint

In the authors' words

We study a mixed regular--singular control problem for stochastic evolution equations in a Hilbert space with possibly unbounded random linear operators, a nonconvex regular-control domain, and a state-dependent singular coefficient. The singular control is an adapted nondecreasing c\`adl\`ag process whose terminal value need not be bounded. We prove well-posedness and weighted moment estimates for the forward and backward equations, and characterize the second-order adjoint by a conditionally expected operator-valued backward stochastic integral equation. An It\^o-type formula for the quadratic form of this adjoint, together with spike and convex variations, yields a second-order Hamiltonian condition for the regular control, as well as nonnegativity and a contact condition for the optional singular Hamiltonian.

Main resultThe abstract does not state a limitation.

Appeared: Friday, October 2. arXiv. Preprint, not yet peer-reviewed.