Novel approaches to the optimal control of complex random systems

Many Operational Research problems concern the control of some complex system which evolves at random. Standard approaches to the development of optimal controls use a methodology called dynamic programming. Sadly, this is computationally infeasible for many systems of realistic size. Current research is geared to the development of efficient alternatives which are close to optimal and which provide interpretable answers.

For further information contact Kevin GlazebrookChris Kirkbride or Peter Jacko.

Return to the Optimisation Research Group page.