Necessary conditions for optimal control of stochastic evolution equations in Hilbert spaces (Q538476)

The paper studies a system governed by a stochastic evolution equation \[ dX(t) = \big(A(t)X(t) + F(X(t),\nu(t))\big)dt + G(X(t))dM(t) \] in a Hilbert space, where \(A(t)\) is an unbounded linear operator, \(F\) and \(G\) are differentiable functions with bonded derivatives, \(M\) is a continuous martingale, and \(\nu(t)\) is a control. The main problem considered in the article is minimizing the cost functional over a set of admissible controls. This problem is approached through using the theory of backward stochastic differential equations for deriving a stochastic maximum principle for this control problem. In fact, the adjoint equation the derived in the paper turns out to be a backward stochastic partial differential equation and it can be dealt with by using previous results by the same author.