论文信息 - Analysis of Complete Convergence for Genetic Algorithm with Immune Memory

Analysis of Complete Convergence for Genetic Algorithm with Immune Memory

A new Immune Memory Genetic Algorithm (IMGA) based on the mechanism of immune memory and immune network is proposed in this article . Using Markov chains theory, we proven that NGA(Niche Genetic Algorithms) can't not be complete convergence but IMGA can. The contrast simulation experiments between NGA and IMGA are performed. The experiments results validate the theoretical analysis and testify that IMGA has availability on solving multi-modal optimization problems, with quickly convergence ability and wonderful stability.

[1] Wang Xiu-feng. Adaptive genetic algorithm for multi-peak searching , 2004 .

[2] Günter Rudolph,et al. Convergence analysis of canonical genetic algorithms , 1994, IEEE Trans. Neural Networks.

[3] Jonathan Timmis,et al. Artificial immune systems - a new computational intelligence paradigm , 2002 .

[4] David E. Goldberg,et al. Genetic Algorithms with Sharing for Multimodalfunction Optimization , 1987, ICGA.

[5] Jonathan Timmis,et al. Artificial Immune Systems: A New Computational Intelligence Approach , 2003 .

[6] John H. Holland,et al. Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .