Minisymposium 1: Domain Decomposition Methods for Simulation-constrained Optimization

Abstract

By simulation we refer to numerical solution of systems governed by partial differential (or integral) equations. Tremendous strides in large-scale algorithms and hardware have provided the framework for high fidelity simulations, to the point that it is now practical to consider complex optimization problems. In such problems we wish to determine various parameters that typically consist the data of a simulation: boundary and initial conditions, material properties, distributed forces, or shape.