Open AccessAN ADAPTIVE CONSTRAINT HANDLING TECHNIQUE FOR PARTICLE SWARM IN CONSTRAINED OPTIMIZATION PROBLEMS
Author(s) -
Érica da Costa Reis Carvalho,
José Pedro Gonçalves Carvalho,
Heder S. Bernardino,
Patrícia H. Hallak,
Afonso C. C. Lemonge
Publication year - 2016
Publication title -
revista ciatec-upf
Language(s) - English
Resource type - Journals
ISSN - 2176-4565
DOI - 10.5335/ciatec.v8i1.6023
Subject(s) - multi swarm optimization , mathematical optimization , particle swarm optimization , penalty method , computer science , benchmark (surveying) , metaheuristic , heuristics , optimization problem , constrained optimization , constraint (computer aided design) , test functions for optimization , constrained optimization problem , derivative free optimization , mathematics , geometry , geodesy , geography
Nature inspired meta-heuristics are largely used to solve optimization problems. However, these techniques should be adapted when solving constrained optimization problems, which are common in real world situations. Here an adaptive penalty approach (called Adaptive Penalty Method, APM) is combined with a particle swarm optimization (PSO) technique to solve constrained optimization problems. This approach is analyzed using a benchmark of test-problems and 5 mechanical engineering problems. Moreover, three variants of APM are considered in the computational experiments. Comparison results show that the proposed algorithm obtains a promising performance on the majority of the test problems