提出一种克隆选择算法——基于等级变异的克隆选择算法.为提高进化中变异的有效性,算法将变异尺度分成若干等级,低等级变异有利于跳出局部最优解,实现全局寻优;高等级变异有利于局部的高精度寻优.此外,算法在进化过程中记忆父抗体的变异尺度等级等信息,并制定有效的变异策略运用这些信息以指导后续进化过程.采用标准函数测试并与其它优化算法进行对比.实验结果表明,该算法具有收敛速度快、全局搜索能力强、精度高和鲁棒性好的优点. In order to improve the validity of mutation in evolution, the algorithm divides the variation scale into several levels, and low-level variation is good for jumping out of the local optimal solution and achieving global optimization. In addition, the algorithm memorizes the parental antibody variation scale and other information in the evolutionary process, and formulates an effective mutation strategy to use these information to guide the subsequent evolution process.With the standard function test and with other Optimization algorithm.Experimental results show that the algorithm has the advantages of fast convergence, global search ability, high precision and good robustness.