Estimation of robustness for controlled diffusion processes

From MaRDI portal





Consider the problem of controlling a system described by a stochastic differential equation whose coefficients are known only up to a certain degree of accuracy. One could choose a suitable system within the range -- for instance, a linear system or one that is smooth enough -- find an \(\varepsilon\)-optimal control policy for this system, and use it for controlling the original system. To what degree is the performance of that control still good in the original system? The robustness inequality for an optimization problem with constraints given by contractive operators is adapted to controlled stochastic differential equations, and applications to the estimation of the approximation accuracy of controlled processes are discussed.











This page was built for publication: Estimation of robustness for controlled diffusion processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4248572)