Augmented Lagrangian algorithms in an infinite-dimensional setting

Project: Research

Project Details

Description

We develop a class of augmented Lagrangian algorithms
for the solution of optimization problems with equality constraints
posed in an infinite-dimensional setting. This research is motivated
by the numerical solution of optimal control problems. The purpose is
to extend, from finite dimensions, a class of augmented Lagrangian
algorithms similar to that implemented in the software
LANCELOT.
StatusActive
Effective start/end date1/09/96 → …

Keywords

  • nonlinear optimization
  • constrained optimization
  • infinite-dimensional setting