Mixed Discontinuous Galerkin Time-Stepping Method for Linear Parabolic Optimal Control Problems

Tianliang Hou,Yanping Chen
DOI: https://doi.org/10.4208/jcm.1211-m4267
2015-01-01
Journal of Computational Mathematics
Abstract:In this paper, we discuss the mixed discontinuous Galerkin (DG) finite element approximation to linear parabolic optimal control problems. For the state variables and the co-state variables, the discontinuous finite element method is used for the time discretization and the Raviart-Thomas mixed finite element method is used for the space discretization. We do not discretize the space of admissible control but implicitly utilize the relation between co-state and control for the discretization of the control. We derive a priori error estimates for the lowest order mixed DG finite element approximation. Moveover, for the element of arbitrary order in space and time, we derive a posteriori L 2 (0,T;L 2 ()) error estimates for the scalar functions, assuming that only the underlying mesh is static. Finally, we present an example to confirm the theoretical result on a priori error estimates.
What problem does this paper attempt to address?