Optimal control of a heat conduction problem using its low order approximation

Abstract

Optimal control of a large dynamical system is accomplished by designing the control strategy on its low order approximation. The Large system is the Finite Element (FE) model of a heat conduction problem and its low order approximation is obtained using Krylov Subspace Projection. It is seen that this approach provides good dividends as the desired cost functional is minimized reasonably at substantially reduced computational cost. A study of the sub-optimality caused is however required and is pointed out as a subject of possible exploration.