引用本文:杨海东,鄂加强.自适应变尺度混沌免疫优化算法及其应用[J].控制理论与应用,2009,26(10):1069~1074.[点击复制]
YANG Hai-dong,E Jia-qiang.An adaptive chaos immune optimization algorithm with mutative scale and its application[J].Control Theory and Technology,2009,26(10):1069~1074.[点击复制]
自适应变尺度混沌免疫优化算法及其应用
An adaptive chaos immune optimization algorithm with mutative scale and its application
摘要点击 1955  全文点击 1480  投稿时间:2008-06-06  修订日期:2008-12-21
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/j.issn.1000-8152.2009.10.CCTA080581
  2009,26(10):1069-1074
中文关键词  混沌免疫优化算法  混沌  免疫  移动Ad Hoc网络
英文关键词  chaos immune optimization algorithm  chaos  immune  mobile Ad Hoc networks
基金项目  国家自然科学基金资助项目(60973132); 广东省自然科学基金资助项目(8451064101000630); 教育部高校博士点基金资助项目(20070561081); 广东省工业科技攻关计划资助项目(2007B010200046).
作者单位E-mail
杨海东* 华南理工大学 自动化科学与工程学院 hdyang@scut.edu.cn 
鄂加强 湖南大学 机械与运载工程学院  
中文摘要
      结合混沌优化算法与免疫算法的特点, 提出了一种采用折叠次数无限的自映射x = sin(2/x)产生混沌变量的自适应变尺度混沌免疫优化算法. 该算法通过自适应变尺度方法不断调整优化变量的搜索空间, 同时采用最大循环次数作为控制指标, 既保证了寻优的准确性, 又保证了算法的快速性. 应用该算法对3个测试函数进行优化计算得到了比较满意的结果. 将此算法应用于移动Ad Hoc网络入侵检测时的仿真实验结果表明, 自适应变尺度混沌 免疫优化算法能有效地减少对训练样本的依赖, 同时减少噪音数据对入侵检测系统性能的影响, 适用于移动自组网络对于入侵检测系统高检测率、高抗噪能力和低计算延迟的要求.
英文摘要
      By combing the chaos optimization method and the immune algorithm, we propose an adaptive chaos immune optimization algorithm(AMSCIOA) with mutative scale, using one-dimensional iterative chaotic self mapping x = sin(2/x) with infinite collapses within the finite region [–1, 1]. In the optimization process, to ensure the high speed and precision some measures are taken, including: 1) the ranges of optimized variables are reduced continuously by the adaptive mutative scale method, and the searching precision is enhanced accordingly; 2) the maximal number of repetitions is regarded as a controlled index. The simulation results for three testing functions validate the high speed and precision of the AMSCIOA with mutative scale. The simulation of the intrusion detection system for detecting the intrusions to mobile Ad Hoc networks show that this algorithm lowers the dependence of training samples, reduces the noise influence on the performance, provides a high detection rate, and produces a small time-delay caused by computation.