Convergence Rate Analysis of a Proposed Function Space Algorithm (FSA)

Abstract

In this work, the numerical implementation of Function Space Algorithms (FSA) for the solution of quadratic continuous cost functional was considered. It is used to solve Reaction Diffusion Control problems. It considered specifically a parabolic problem characterized by dynamics constraints and the results obtained analyzed. The cumbersome nature of the line search techniques associated with FSA was addressed by time discretization approach. It is shown that the convergence rate of FSA improves as the penalty parameter grows.