An improved artificial immune algorithm for multimodal optimization

作者：Xie Jing Xin^*; Cheng Chun Tian; Tong Lei Guang

来源：Journal of Harbin Institute of Technology, 2009, 41(7): 135-139.

• Affinity
• Artificial immune algorithms
• Average entropy
• Benchmark functions
• Binary encodings
• Concentration values
• Euclidean distance measure
• Multi modal function
• Multi-modal optimization
• Multimodal function optimization
• Multimode function
• Numerical experiments
• Real values

To improve the efficiency of basic artificial immune algorithm (AIA), this paper presents an improved artificial immune algorithm (IAIA) for multimodal function optimization. Appropriate cluster analysis was conducted on real values of individuals with Euclidean distance measure, then all individuals within a cluster were considered as a group with the same concentration value. The average entropy on the whole group was calculated with binary encoding of individuals, which replaced the conventionally calculation of entropy between two individuals in AIA. To maintain high diversity, the fitness of each individual and concentration were taken into account in determining reproduction probability. Two benchmark functions were used to demonstrate the validity of IAIA and the role of each design of IAIA. Numerical experiments show that IAIA can reduce the complexity of computation, and then increases the efficiency of AIA with the maintenance of diversity and convergence in optimizing multimodal functions.

• 出版日期2009