Advanced search
Start date
Betweenand

Development of a hybrid metaheuristic with adaptive control flow and parameters

Grant number: 18/15417-8
Support type:Research Grants - Young Investigators Grants- Phase 2
Duration: May 01, 2019 - April 30, 2024
Field of knowledge:Engineering - Production Engineering
Principal Investigator:Antônio Augusto Chaves
Grantee:Antônio Augusto Chaves
Home Institution: Instituto de Ciência e Tecnologia (ICT). Universidade Federal de São Paulo (UNIFESP). Campus São José dos Campos. São José dos Campos , SP, Brazil
Assoc. researchers:Luiz Antonio Nogueira Lorena
Associated research grant:12/17523-3 - New hybrid methods to resolve combinatorial optimization problems, AP.JP

Abstract

The study of efficient metaheuristics to solve optimization problems has been the subject of much research by the scientific community. To obtain good results in terms of solution quality and computational time it is important to have a good configuration of the metaheuristic. This process of specifying control flow and parameter values of a method is a hard task. Thus, this project has as main idea the development and improvement of the adaptive Biased Random-key Genetic Algorithm (A-BRKGA) method to choose which components will be used and in which sequence (A-BRKGA flow) and which parameters to use while an instance of a problem is being solved. To this end, machine learning techniques and adaptive and reactive mechanisms will be studied to construct an A-BRKGA with online configuration of parameters and control flow. The goal is to generate an efficient algorithm to solve combinatorial optimization problems and make the code easy to reuse. In order to evaluate the proposed method, four optimization problems with industrial and logistical applications will be studied: field technician scheduling problem, multicommodity traveling salesman problem with priority prizes, two-stage capacitated facility location problem, and facility location problem with overlapping. The computational tests will use test problems available in the literature and real case studies. The method will be compared with state-of-the-art algorithms through statistical analysis. (AU)

Scientific publications
(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)
DA SILVA, TIAGO TIBURCIO; CHAVES, ANTONIO AUGUSTO; YANASSE, HORACIO HIDEKI; LOUREIRO LUNA, HENRIQUE PACCA. The multicommodity traveling salesman problem with priority prizes: a mathematical model and metaheuristics. COMPUTATIONAL & APPLIED MATHEMATICS, v. 38, n. 4 DEC 2019. Web of Science Citations: 0.

Please report errors in scientific publications list by writing to: cdi@fapesp.br.