Rank-density-based multiobjective genetic algorithm and benchmark test function study

doi:10.1109/TEVC.2003.812220

Return

Rank-density-based multiobjective genetic algorithm and benchmark test function study

2003-08-01

158

PRE

Haiming Lu

Gary G. Yen

DOI：10.1109/TEVC.2003.812220

Original

Original request for help

Save

Abstract

En 中文

Since the 1980s, the application of evolutionary algorithms (EAs) in solving multiobjective optimization problems (MOPs) has been receiving a growing interest from evolutionary computation community. To search for a family of acceptable solutions, a so called Pareto set, by using EAs population-based parallel searching ability, several multiobjective evolutionary algorithms (MOEAs) have been proposed. However, most of these MOEAs have difficulty in dealing with the tradeoff between uniformly distributing the computational resources and finding the near-complete and near-optimal Pareto set. On the other hand, according to the no-free-lunch theorems, no formal assurance of an algorithm's general effectiveness exists if insufficient knowledge of the problem characteristics is incorporated into the algorithm domain. In this paper, we propose a new evolutionary approach to MOPS, the rank-density-based genetic algorithm (RDGA) that synergistically. integrates selected features from existing MOEAs in a unique way. A new ranking method, automatic accumulated ranking strategy, and a forbidden region concept are introduced, completed by a revised adaptive cell density evaluation scheme and a rank-density-based fitness assignment technique. In addition, four types of MOP features, such as discontinuous and concave Pareto front, local optimality, high-dimensional decision space and high-dimensional objective space are exploited and the corresponding MOP test functions are designed. By examining the selected performance indicators, RDGA is found to be statistically competitive with four state-of-the-art MOEAs in terms of keeping the diversity of the individuals along the tradeoff surface, tending to extend the Pareto front to new areas. and finding a well-approximated Pareto optimal front.

Keywords:

multiobjective evolutionary algorithm (MOEA)

multiobjective optimization (MO)

Pareto optimality