Space-time residual minimization with efficient linear solver based on matrix compression

Marcin Loś

AGH University of Science and Technology in Kraków, Poland.


We present a space-time formulation of the non-stationary advection-dominated advection-diffusion problem based on a constrained least-squares approach. For discretization, we use Isogeometric Analysis and employ B-spline basis functions. While using standard separate time and space discretization, in some restricted cases, it is possible to construct highly efficient linear solvers exploiting the structure of the problem and the time stepping schemes, with space-time formulation, the resulting matrix has a more complex structure. We present the idea of an iterative solver employing a matrix compression technique based on recursive decomposition and singular value decomposition (SVD).




We will meet in Google Meet, use the link below to connect with us

September 30, 2022, 15:00 hrs. Chile, room IMA 2-3 (IMA-PUCV) and https://meet.google.com/viw-rqds-ikc

If you are interested in giving a talk, please contact: patrick.vega@pucv.cl