scientific article; zbMATH DE number 3932148
The author obtains a closed-form expression for the minimum value of the expected cost in a stochastic control problem where the state equation is of diffusion type, linear in control and the cost function is quadratic in control, but otherwise nonlinear in the state. More precisely, let the state \(\xi_ t\), \(t\geq 0\), and expected cost J(u), be given, respectively, by \[ d\xi_ t=u(t,\xi)dt+dW_ t,\quad \xi_ 0=x\in R^ d,\quad J(u)=E[I(\xi)+1/2\int^{\tau_{\xi}}_{0}| u(t,\xi)|^ 2dt] \] where \(\tau_{\xi}\) is a stopping time and I is an integrable function bounded from below. With u belonging to a certain class \({\mathfrak U}\) of nonanticipative control functions, the main result of the paper is \[ \min_{u\in {\mathfrak U}}J(u)=-\log E^ 0[\exp [-I(y)]] \] where \(E^ 0\) is the expectation with respect to a new measure on the space of d-dimensional continuous functions. The proof of the theorem utilizes decomposition methods and the theory of representation of martingales.
- Source intensity control in transport and diffusion problems
- A new actuator activation policy for performance enhancement of controlled diffusion processes.
- Control of heterogeneous diffusion process in order to stabilize the probability distribution density
- Control Theory and Experimental Design in Diffusion Processes
- scientific article; zbMATH DE number 4192502 (Why is no real title available?)
- CONTROLLED DIFFUSION PROCESSES WITE SUCCESSIVE REWARDS
- scientific article; zbMATH DE number 4006128 (Why is no real title available?)
- scientific article; zbMATH DE number 4003177 (Why is no real title available?)
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3704699)