基于人工鱼群算法的网络流量预测方法  

Network traffic forecasting method based on artificial fish swarm algorithm

作　　者：马浩[1] MA Hao(Department of Computer Science & Technology,Yuncheng University,Yuncheng 044000,China)

机构地区：[1]运城学院计算机科学与技术系,山西运城044000

出　　处：《现代电子技术》2016年第18期73-75,79,共4页Modern Electronics Technique

基　　金：国家自然科学基金项目(61303232);2015年广东省佛山市机电专业群工程技术开发中心开放基金:基于物联网技术的产品质量控制系统设计与开发(2015-KJZX069)

摘　　要：使用非线性预测模型支持向量回归算法建立预测模型,对具有明显非平稳性、混沌性以及非线性的网络流量进行预测研究。使用人工鱼群算法对支持向量回归算法的参数进行寻优,使用PSO算法对常规人工鱼群算法进行改进,使得人工鱼不依赖步长因子,仅对视野因子产生依赖,能够得到最优解。通过使用Logistic映射对人工鱼位置进行初始化,提高种群多样性,从而提高算法全局寻优能力,避免算法陷入局部最小值。最后使用MAWI数据集中的三组不同时间粒度的数据进行网络预测方法的实例分析,结果表明,所研究的人工鱼群算法具有较好的预测性能,能够满足网络流量预测的需求。The support vector regression algorithm of the nonlinear prediction model is used in this paper to establish theforecasting model to study the prediction of the network traffic with obvious nonstationarity,chaos and nonlinearity. In this pa?per,the artificial fish swarm algorithm is adopted to optimize the parameters of support vector regression algorithm. The PSO al?gorithm is used to improve the conventional artificial fish swarm algorithm. The Logistic map is used to initialize the position ofthe artificial fish to improve the diversity of the population,so as to improve the global optimization ability of the algorithm andavoid the algorithm falling into local minimum value. The data of three groups with different time granularity are analyzed byusing MAWI data. The results show that the artificial fish swarm algorithm has better prediction performance and can meet theneeds of network traffic prediction.

关 键 词：网络流量预测 人工鱼群算法 支持向量回归 混沌机制 粒子群优化 

分 类 号：TN926.34[电子电信—通信与信息系统] TP393[电子电信—信息与通信工程]