A MODIFIED IMMUNE OPTIMIZATION ALGORITHM
(0) This paper proposes a modified immune optimization algoritom for the multi-modal problems. It is inspired by the clonal selection and antibody diversity maintaining principles in the immunology. Compared with the existing immune optimization algorithms, our new algorithm defines the concentration of the antibody to stand for the solution diversity, and the mutation rate is dynamically adjusted based on the antibody concentration and fitness. Moreover, this algorithm takes advantage of the clusters of antibodies. For each cluster,the elitist ones are always retained as the memory set. Therefore, an improved trade-off between the exploration and exploitation in the solution space can be achieved. A few multi-modal and high-dimension benchmark functions are utilized here to examine the efficiency of our optimization method. Performance comparisons are also made with the CLONALG and opt-aiNET. Simulation results demonstrate that this modified optimization algorithm can effectively ob tain the optimal solutions, and still maintain the solution diversity, which is crucial for dealing with challenging real-world optimization problems.
Immune optimization algorithm Diversity Cional selection Multi-modal optimization Mutation rate
ZHUO-YUE SONG X.Z.GAO XIAN-LIN HUANG H.S.LIN
Center for Control Theory and Guidance Technology, Harbin Institute of Technology, Harbin, China Institute of Intelligent Power Electronics, Helsinki University of Technology, Espoo, Finland
国际会议
2006 International Conference on Machine Learning and Cybernetics(IEEE第五届机器学习与控制论坛)
大连
英文
2184-2189
2006-08-13（万方平台首次上网日期，不代表论文的发表时间）