Authors:
Guang Peng
;
Zhihao Shang
and
Katinka Wolter
Affiliation:
Department of Mathematics and Computer Science, Free University of Berlin, Takustr. 9, Berlin and Germany
Keyword(s):
Multiobjective Optimization, Evolutionary Computation, Decomposition, Artificial Bee Colony, Adaptive Normalization.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Computational Intelligence
;
Evolutionary Computing
;
Soft Computing
;
Swarm/Collective Intelligence
Abstract:
This paper presents a multiobjective artificial bee colony (ABC) algorithm using the decomposition approach for improving the performance of MOEA/D (multiobjective evolutionary algorithm based on decomposition). Using a novel reproduction operator inspired by ABC, we propose MOEA/D-ABC, a new version of MOEA/D. Then, a modified Tchebycheff approach is adopted to achieve higher diversity of the solutions. Further, an adaptive normalization operator can be incorporated into MOEA/D-ABC to solve the differently scaled problems. The proposed MOEA/D-ABC is compared to several state-of-the-art algorithms on two well-known test suites. The experimental results show that MOEA/D-ABC exhibits better convergence and diversity than other MOEA/D algorithms on most instances.