857. The novel interior augmented Lagrangian methods for constrained optimization
Invited abstract in session WD-35: Bilevel optimization and augmented Lagrangian methods, stream Continuous and mixed-integer nonlinear programming: theory and algorithms.
Wednesday, 14:30-16:00Room: Michael Sadler LG15
Authors (first author is the speaker)
| 1. | Xinwei Liu
|
| Hebei University of Technology | |
| 2. | Yu-Hong Dai
|
| ICMSEC, AMSS,Chinese Academy of Sciences |
Abstract
Interior-point approach is one kind of most important and effective methods for nonlinear inequality-constrained optimization, whereas classic augmented Lagrangian methods are very efficient in solving nonlinear equality-constrained optimization and convex optimization. The newly research has shown that the interior-point augmented Lagrangian function can be of higher-order smoothness, and is convex when the original optimization problem is convex. These nice properties make the novel augmented Lagrangian applicable and effective not only for convex optimization but also for non-convex optimization problems. We summarize our interior-point augmented Lagrangian methods on non-convex and convex optimization, and point out some next research topics.
Keywords
- Mathematical Programming
- Interior Point Methods
- Continuous Optimization
Status: accepted
Back to the list of papers