Convergence of a conditional gradient algorithm for relaxed controls involving first boundary value problems of parabolic type

Abstract

In this paper, we consider a class of optimal control problems involving linear first boundary value problems of parabolic type. A conditional gradient method is used to devise a computational algorithm for solving this class of problems. The convergence property of this algorithm is investigated by using a topology arising in the study of relaxed controls.

Keywords:

• Optimal control
• parabolic partial differential equations
• first boundary conditions
• conditional gradient algorithm
• convergence
• relaxed control

¹This work was partially supported by the Australian Research Grants Committee.

¹This work was partially supported by the Australian Research Grants Committee.

Notes

¹This work was partially supported by the Australian Research Grants Committee.