Integrating nonlinear branch-and-bound and outer approximation for convex Mixed Integer Nonlinear Programming

In this paper, we present a new hybrid algorithm for convex Mixed Integer Nonlinear Programming (MINLP). The proposed hybrid algorithm is an improved version of the classical nonlinear branch-and-bound (BB) procedure, where the enhancements are obtained with the application of the outer approximatio...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of global optimization Jg. 60; H. 2; S. 373 - 389
Hauptverfasser: Melo, Wendel, Fampa, Marcia, Raupp, Fernanda
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Boston Springer US 01.10.2014
Springer
Schlagworte:
Algorithms
Comparative analysis
Computer Science
Mathematics
Mathematics and Statistics
Operations Research/Decision Theory
Optimization
Real Functions
Branch-and-bound
Mixed Integer Nonlinear Programming
Outer approximation
Hybrid algorithm
ISSN:0925-5001, 1573-2916
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this paper, we present a new hybrid algorithm for convex Mixed Integer Nonlinear Programming (MINLP). The proposed hybrid algorithm is an improved version of the classical nonlinear branch-and-bound (BB) procedure, where the enhancements are obtained with the application of the outer approximation algorithm on some nodes of the enumeration tree. The two methods are combined in such a way that each one collaborates to the convergence of the other. Computational experiments with benchmark instances of the MINLP problem show the good performance of the proposed algorithm, which is compared to the outer approximation algorithm, the nonlinear BB algorithm and the hybrid algorithm implemented in the solver Bonmin.
ISSN:0925-5001
1573-2916
DOI:10.1007/s10898-014-0217-8