航空网络关键节点辨识的核极限学习机算法研究  

Research on Kernel Extreme Learning Machine Algorithm for Key Node Identification in Aviation Network

作　　者：牛军锋[1] 甘旭升[2] 孙静娟[2] 涂从良 NIU Junfeng;GAN Xusheng;SUN Jingjuan;TU Congliang(Department of Management Technology,Xijing University,Xi’an 710123,China;College of Air Traffic Control and Navigation,Air Force Engineering University,Xi’an 710051,China)

机构地区：[1]西京学院管理技术系,西安710123 [2]空军工程大学空管领航学院,西安710051

出　　处：《航空工程进展》2021年第1期39-47,共9页Advances in Aeronautical Science and Engineering

摘　　要：通过技术手段准确识别航空网络的关键节点,对航空网络平时的正常运行以及战时的防御和修复,具有重要的理论意义和参考价值。提出一种基于核极限学习机的航空网络关键节点识别方法,首先,采用层次分析法对节点综合重要度进行评估;然后,选取三个简单指标,基于核极限学习机学习简单指标与综合重要度之间的映射关系,建立重要度评估模型;最后,以中美两国航空网络为例进行仿真。结果表明:仅需计算40个节点的复杂指标值,就可对关键节点取得较满意的辨识效果,降低了计算复杂度,提高了辨识效率,即采用本文方法辨识航空网络的关键节点是有效、可行的。Accurately identifying the key nodes of aviation network through technical means is of important theoretical significance and reference value for the normal operation of aviation network in peacetime and defense and repair in wartime.For this,a key node identification method based on kernel extreme learning machine is proposed.Firstly,the comprehensive importance of nodes based on analytic hierarchy process(AHP)is evaluated.Then,three simple indices are selected and the importance evaluation model is established based on the mapping relationship between simple indices and comprehensive importance of kernel extreme learning machine.Finally,the simulation is carried out with the example of China-US air network.The simulation analysis of the American aviation network indicates that it can obtain satisfactory identification effect for key nodes by only calculating the complex index value of 40 nodes,which reduces the calculation complexity and improves the identification efficiency.This shows that it is effective and feasible to use this method to identify the key nodes of aviation network.

关 键 词：航空网络 关键节点辨识 极限学习机 参数寻优 核函数 

分 类 号：V35[航空宇航科学与技术—人机与环境工程]