一种新的免疫进化算法及其性能分析

基于免疫系统中的进化机理，提出了一种免疫进化算法。首先引入了邻域概念，并通过定义扩展半径和突变半径两个新算法参数而构造了较小和较大两个邻域。进而给出了扩展和突变操作分别利用这两个邻域进行局部和全局搜索，实现了从全局到局部的两层邻域搜索机制。分析了算法的优化机理和收敛性。仿真结果表明该算法具有不易陷入局部最优、解的精度高、收敛速度快等优点。

A New Immune Evolutionary Algorithm and Its Performance Analysis

An Immune Evolutionary Algorithm is proposed based on evolution principle in the immune system. First, the concept of neighborhood is introduced, and two new algorithm parameters of expansion radius and mutation radius are defined to construct a small neighborhood and a large neighborhood. Second, expansion and mutation operations are given, which use the two neighborhoods to perform local and global search respectively, so that two-level neighborhood search mechanism is realized that searches from global to local regions. The optimization principle and convergence property of the algorithm are analyzed. The simulation results show that the algorithm has nice performances such as avoiding local optima, high precision solution, and quick convergence.