这是一个非常简单的遗传算法源代码，是由Denis Cormier (North Carolina State University)开发的，Sita S.Raghavan (University of North Carolina at Charlotte)修正。代码保证尽可能少，实际上也不必查错。对一特定的应用修正此代码，用户只需改变常数的定义并且定义“评价函数”即可。注意代码 的设计是求最大值，其中的目标函数只能取正值；且函数值和个体的适应值之间没有区别。该系统使用比率选择、精华模型、单点杂交和均匀变异。如果用 Gaussian变异替换均匀变异，可能得到更好的效果。代码没有任何图形，甚至也没有屏幕输出，主要是保证在平台之间的高可移植性。读者可以从ftp.uncc.edu, 目录 coe/evol中的文件prog.c中获得。要求输入的文件应该命名为‘gadata.txt’；系统产生的输出文件为‘galog.txt’。输入的 文件由几行组成：数目对应于变量数。且每一行提供次序——对应于变量的上下界。如第一行为第一个变量提供上下界，第二行为第二个变量提供上下界，等等。 -This is a very simple genetic algorithm is Denis Cormier source of Carolina re (State) development, Sita leads S.R aghavan (of Carolina at leads, where re. Code that is actually less as far as possible, don t find fault. For a particular application of this code, the user need revision of the constant change and define "evaluation function definition of". Note the design code for maximum, which is the objective function can take positive, And the function of the individual value and no difference between fitness. This system USES ratio, essence model, single hybridization and uniform variation. If use uniform variation and variation of Gaussian replacement may get better effect. Code without any graphics, nor even screen output, mainly is the guarantee of the platform between high portability. Readers can from the FTP uncc. J, directory coe/evol edu files in prog. C. The documents required input should be named "j"　rather gadata System to produce output file for galog. J TXT　 Input f