中国民航大学学报 ›› 2019, Vol. 37 ›› Issue (4): 41-45.

• 民用航空 • 上一篇    下一篇

基于AHP和D-S证据理论的民航信息系统风险评估#br#

张西武1,李志平2a,李跃凯2b
  

  1. (1.中国民用航空局信息中心,北京100710; 2.中国民航大学a.计算机学院; b.信息网络测评中心,天津300300)
  • 出版日期:2019-08-23 发布日期:2020-04-01
  • 作者简介:张西武(1970—)男,安徽安庆人,硕士,研究方向为民航网络安全管理.
  • 基金资助:
    民航安全能力建设资金项目(PESA0001, PE2018052)

Risk assessment of civil aviation information system based on AHP and D-S evidence theory#br#

ZHANG Xiwu1, LI Zhiping2a, LI Yuekai2b#br#   

  1. (1. Information Center of CAAC, Beijing 100710, China; 2a. College of Computer Science and Technology;2b. ISECCA, CAUC, Tianjin 300300, China)
  • Online:2019-08-23 Published:2020-04-01

摘要: 针对传统民航信息系统安全评估周期长、过程繁琐及专家主观性较强等问题,提出一种基于层次分析法和改进Dempster-Shafer(D-S)证据理论的风险评估方法。首先基于层次分析法确定各层因素相对于上层因素的权重, 利用模糊综合评价给出最底层因素的评语集隶属度矩阵; 然后利用基于矩阵分析法和信息熵的D-S 合成算法计算证据间冲突程度K 并对证据重新赋予权值,得出证据对各因素的支持程度;最后通过综合计算来确定系统最终安全等级。该方法进一步降低了专家评估的主观性,客观高效,操作方便,适用于民航信息系统的安全风险评估。

关键词: AHP, D-S 证据理论, 模糊综合评价, 隶属度矩阵, 信息熵

Abstract: Aim at the long period, complex process and strong subjectivity of experts in the traditional risk assessing procedure of civil aviation information system, a method based on AHP and improved D-S evidence theory is proposed. Firstly, the weights of each layer of factors relative to the upper layer are determined basing on AHP,and then a fuzzy comprehensive evaluation is used to give the commentary membership matrix of the lowest level factors; After that, through D-S synthesis algorithm based on matrix analysis and information entropy, the degree of conflict K between evidences is calculated and the evidence is re-weighted, obtaining the supporting degree of evidence for each factor; Final security level of the system is determined by comprehensive calculation.Compared with the widely used traditional fuzzy comprehensive evaluation method, the proposed method further decreases the expert subjectivity and is more efficient and easier to operate for safety risk assessment of civil aviation information system.

Key words: AHP, D-S evidence theory, fuzzy comprehensive evaluation, membership matrix, information entropy

中图分类号: