代谢灵敏度分析在网络故障诊断与自愈中的应用  

Application of Network Fault Diagnosis and Self-healing Based on Metabolic Sensitivity Analysis

作　　者：胡扬[1] 桂卫华[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《湖南大学学报（自然科学版）》2009年第6期37-40,共4页Journal of Hunan University:Natural Sciences

基　　金：国家自然科学基金资助项目(60874008);湖南省自然科学基金资助项目(07JJ6126)

摘　　要：以交通运输网络为例对网络化系统进行了建模分析,利用人工代谢系统的代谢灵敏度折线图对网络支路流量变化进行了分析,得到了故障产生的瓶颈位置.通过调节代谢算子对网络拥塞程度进行了控制,使网络流量趋于平衡,最终实现网络自愈.实例分析表明:人工代谢系统能实现故障的准确定位,代谢反馈调节机制在网络自愈中起到了关键作用.Fault diagnosis is a complex problem in network system. The main reason is that there may be some possible factors in the fault. Taking a transportation network for example, the network system was modeled and analyzed. The situation of traffic flow in network paths was analyzed by metabolic sensitivity curve in artificial metabolic system and the fault bottleneck was gained. Network congestion could be controlled by adjusting the metabolic operators, and load balance in network traffic flow could be realized. At last, the aim of self-healing for the total network system was reached. The fault points could be located accurately with an artificial metabolic system. The artificial metabolism feedback mechanism played a key role in network self- healing.

关 键 词：新陈代谢 人工代谢系统 代谢灵敏度 故障诊断 网络自愈 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]