A review of Simheuristics: extending metaheuristics to deal with stochastic optimization problems

• Type: Article in International Scientific Journal
• Date: 2015

Authors

Gonçalo Figueira (Author) FEUP; Angel A. Juan (Author) Other; Javier Faulin (Author) Other; Scott E. Grasman (Author) Other; Markus Rabe (Author) Other

Journal

Title: Operations Research Perspectives

Vol. 2

Pages: 62-72

ISSN: 2214-7160

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 155 Citations

Scopus - 223 Citations

Scientific classification

FOS: Engineering and technology

Other information

Authenticus ID: P-00G-3DK

DOI: 10.1016/j.orp.2015.03.001

Abstract (EN): Many combinatorial optimization problems (COPs) encountered in real-world logistics, transportation, production, healthcare, financial, telecommunication, and computing applications are NP-hard in nature. These real-life COPs are frequently characterized by their large-scale sizes and the need for obtaining high-quality solutions in short computing times, thus requiring the use of metaheuristic algorithms. Metaheuristics benefit from different random-search and parallelization paradigms, but they frequently assume that the problem inputs, the underlying objective function, and the set of optimization constraints are deterministic. However, uncertainty is all around us, which often makes deterministic models oversimplified versions of real-life systems. After completing an extensive review of related work, this paper describes a general methodology that allows for extending metaheuristics through simulation to solve stochastic COPs. ‘Simheuristics’ allow modelers for dealing with real-life uncertainty in a natural way by integrating simulation (in any of its variants) into a metaheuristic-driven framework. These optimization-driven algorithms rely on the fact that efficient metaheuristics already exist for the deterministic version of the corresponding COP. Simheuristics also facilitate the introduction of risk and/or reliability analysis criteria during the assessment of alternative high-quality solutions to stochastic COPs. Several examples of applications in different fields illustrate the potential of the proposed methodology.

Language: English

Type (Professor's evaluation): Scientific