INVESTIGADORES
PILOTTA Elvio Angel
capítulos de libros
Título:
Inexact-Restoration methods for nonlinear programming: Advances and Perspectives
Autor/es:
JOSÉ MARIO MARTÍNEZ; ELVIO A. PILOTTA
Libro:
Optimization and Control with Applications. (Applied Optimization Series, V. 96)
Editorial:
Springer
Referencias:
Lugar: New York; Año: 2005; p. 271 - 292
Resumen:
Abstract.Inexact Restoration methods have been introduced in the last few yearsfor solving nonlinear programming problems. These methods are relatedto classical restoration algorithms but also have some remarkabledifferences. They generate a sequence of generally infeasibleiterates with intermediate iterations that consist of inexactlyrestored points. The convergence theory allows one to use arbitraryalgorithms for performing the restoration. This feature is appealingbecause it allows one to use the structure of the problem in quiteopportunistic ways. Different Inexact Restoration algorithms areavailable. The most recent ones use the trust-region approach.However, unlike the algorithms based on sequential quadraticprogramming, the trust regions are centered not in the current pointbut in the inexactly restored intermediate one. Global convergencehas been proved, based on merit functions of augmented Lagrangiantype. In this survey we point out some applications and we relaterecent advances in the theory.