| Article ID: | iaor20063676 |
| Country: | China |
| Volume: | 17 |
| Issue: | 1 |
| Start Page Number: | 117 |
| End Page Number: | 120 |
| Publication Date: | Jan 2005 |
| Journal: | Chinese Journal of Systems Engineering and Electronics |
| Authors: | Zhang Yi, Yang Xiuxia |
| Keywords: | heuristics |
Based on biological immunology, a new immune genetic algorithm (IGA) is presented, which introduces new immune operators that are realized by vaccination, diversity adjustment of locus and immunelection. The strategies for calculating, judging and adjusting the diversity of population by calculating the locus information entropy are proposed. The effects of control parameters are addressed and the methods of selecting and constructing a vaccine with the system information for TSP are given. The theoretical analysis and the simulation tests on TSP in 75 cities show that IGA can avoid premature convergence of genetic algorithm and improve the convergence speed and precision.