基于变异搜索的人工免疫抗体分布优化研究

Antibody-set Distribution Optimization of Artificial Immune System based on Mutation Searching

优化抗体集合分布以提高抗体集合对非自体空间的覆盖是减少孔洞的重要途径。采用抗体间重叠度量函数作为抗体分布优化的代价函数,研究了基于变异搜索方法的抗体分布优化算法的基本框架,并提出将单次迭代中成功变异的抗体个数作为优化收敛的判定条件之一以减少优化迭代次数。仿真结果表明:抗体分布优化可以大大降低孔洞体积,并且通过监测成功变异抗体的个数可以大大减少迭代次数。

Antibody distribution was optimized in order to improve antibody set coverage in non-self space and reduce the Hole.Taking inner-antibody overlapping function as cost function of antibody distribution optimization,antibody distribution optimization algorithm based on mutation searching was studied,in which the successful mutated antibody number each iteration was regarded as one of termination conditions.Simulation results showed that antibody distribution optimization effectively reduced the volume of Hole,and reduced the iteration number through monitoring successful mutated antibody number.