优化记忆树的研究与实现

The study and realization of utilization memory tree

分析当前人工免疫算法和模型，特别是记忆检测细胞的匹配效率及随着时间推移细胞越来越多占用资源空间的问题．RLAIS模型没有将时间作为控制因子抑制资源的膨胀，不能很好解决随时间推移细胞占用资源越来越多的问题，其一些改进模型在应用中可调节性也不高．为解决此问题而提出一优化记忆树模型．该模型特点：增加时间控制因子；对经常被匹配到的细胞的动态调整．利用时间控制条件和动态调整方法的记忆树模型既优化了记忆细胞的匹配效率，又优化了细胞资源空间，最终实现优化资源空间和提高效率的目的．最后实验验证本模型的可行性．

Some previous artificial immune paradigms and models are analyzed and studied, especially focused on the challenging problems of matching efficiency of memory cells, and the cells occupying more resource spaces with passing of time. RLAIS model doesn＇t take time as a factor to control the expansion of resource, so it can not deal with those cells occupying more resource spaces with passing of time, the regulation of some other optimization models is not very good. With regard to this problem, this paper presents a memory tree model. The most important methods presented in this paper are Time control and Dynamic regulation, which have not been used in the ARB model. The memory tree model with the Time control and Dynamic regulation has not only improved the matching efficiency of memory cells, but also further optimized resources utilization ratio. In the end, real usage data are used to illustrate the working of this novel computational model.