摘要： 基于人工免疫原理，建立了一个基于免疫机制求解TSP问题的数学模型。在该模型中，定义了TSP问题中的抗原和抗体，描述了记忆细胞动态进化过程，并借鉴遗传算法中基因变异思想，提出了优势基因进化的GFE算法，结合生物免疫系统抗体浓度稳定原理，在克隆选择过程中实现了抗体集合的进化计算，快速有效地求解出问题的全局近似最优解。实验结果表明该算法对解决组合优化问题不仅可行，而且有较快的收敛速度和较强的全局搜索能力。

关键词: 人工免疫, 克隆选择, 浓度, 货郎担问题

Abstract: A mathematical model of TSP problem based on artificial immune system is established. In this model, antibody, antigen, and describe dynamic process of memory cells to the problem are defined. This paper refers to the conception of variation GA, and presents GFE algorithm of evolution of superior genes, and makes use of the principle of immune stability in the clone’s procedure to evolve the set of antibody in order to get the global approximate solution rapidly and efficiently. The result of the experiment has proved that the algorithm is not only feasible, but also a comparative good method which has a fast convergence speed and global optimization to solve the NP problems.

Key words: Artificial immune, Clone selection, Concentration, Traveling salesman problem