An Immune Genetic Algorithm for Constrained Multi-modal Optimization Problems

LIAO Min,ZHANG Zhu-hong,WANG Lei

DOI: https://doi.org/10.3969/j.issn.1000-5269.2011.05.018

2011-01-01

Abstract:Firstly transforms the constrained multi-modal optimization problem was transformed into one constrained multi-objective optimization problem by introducing a local detection function.Second,an immune genetic algorithm was proposed to solve such multi-objective problem based on the clonal selection principle and the concept of Pareto dominance.In the design of algorithm,the ideas of non-dominated sorting and niching were utilized to create diverse antibodies.Further,by means of the metaphors of the immune response and gene exchange,several evolving modules were designed to change such antibodies through an island-like evolution fashion so that the optima can be found.Comparative experiments show the proposed approach with satisfactory diversity is potential for constrained multi-modal optimization.