Premium
A chaotic immune algorithm with fuzzy adaptive parameters
Author(s) -
He Hong,
Qian Feng,
Du Wenli
Publication year - 2008
Publication title -
asia‐pacific journal of chemical engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.348
H-Index - 35
eISSN - 1932-2143
pISSN - 1932-2135
DOI - 10.1002/apj.204
Subject(s) - clonal selection , somatic hypermutation , chaotic , premature convergence , adaptability , convergence (economics) , computer science , population , fuzzy logic , mathematical optimization , genetic algorithm , algorithm , artificial intelligence , mathematics , biology , machine learning , antibody , immunology , ecology , demography , b cell , sociology , economics , economic growth
Abstract Adversity loss of the population usually leads to premature convergence problem of the algorithm. To deal with this problem, this paper presents a chaotic hypermutation immune algorithm (CHIA), which is built upon clonal selection principle and immune network regulatory mechanism. In CHIA, a bidirection adaptive chaotic hypermutation is introduced to create antibodies of a new genetic type. Simultaneously, a new immune network regulatory strategy based on the stimulation level of the antibody is devised to maintain current population adversity. Intensive comparison between CHIA and clonal selection algorithm (CLONALG) on function optimization is carried out and the simulation results show that CHIA can overcome the premature convergence problem in CLONALG. Additionally, parameter setting is a crucial and time‐consuming task in the algorithm design. Through the parameter analysis of CHIA, two fuzzy modules are constructed to tune hypermutation factor and death ratio automatically, which eventually form an improved fuzzy adaptive chaotic immune algorithm (FACIA). According to the encouraging results obtained, it can be concluded that FACIA can significantly enhance the adaptability and improve the convergent performance of CHIA through the application of fuzzy logic in parameter setting. Copyright © 2008 Curtin University of Technology and John Wiley & Sons, Ltd.