文章摘要
张宇航,倪明,孙永辉,等.针对网络攻击的配电网信息物理系统风险量化评估[J].电力系统自动化,2019,43(21):12-22. DOI: 10.7500/AEPS20190401003.
ZHANG Yuhang,NI Ming,SUN Yonghui, et al.Quantitative Risk Assessment of Cyber-Physical System for Cyber-attacks in Distribution Network[J].Automation of Electric Power Systems,2019,43(21):12-22. DOI: 10.7500/AEPS20190401003.
针对网络攻击的配电网信息物理系统风险量化评估
Quantitative Risk Assessment of Cyber-Physical System for Cyber-attacks in Distribution Network
DOI:10.7500/AEPS20190401003
关键词: 风险评估  信息物理系统  网络攻击  攻击图
KeyWords: risk assessment  cyber-physical system  cyber-attack  attack graph
上网日期:2019-08-13
基金项目:国家重点研发计划资助项目(2017YFB0903000);国家自然科学基金重点项目(61833008);国家电网公司科技项目“针对网络攻击的电网信息物理系统协同运行态势感知与主动防御方法研究”
作者单位E-mail
张宇航 南瑞集团有限公司(国网电力科学研究院有限公司), 江苏省南京市 211106
智能电网保护和运行控制国家重点实验室, 江苏省南京市 211106
河海大学能源与电气学院, 江苏省南京市 210098 
 
倪明 南瑞集团有限公司(国网电力科学研究院有限公司), 江苏省南京市 211106
智能电网保护和运行控制国家重点实验室, 江苏省南京市 211106
河海大学能源与电气学院, 江苏省南京市 210098 
ni-ming@sgepri.sgcc.com.cn 
孙永辉 河海大学能源与电气学院, 江苏省南京市 210098  
李满礼 南瑞集团有限公司(国网电力科学研究院有限公司), 江苏省南京市 211106
智能电网保护和运行控制国家重点实验室, 江苏省南京市 211106 
 
摘要:
      现代物理电力系统的可靠和安全运行依赖于相关的信息系统,针对信息系统的网络攻击会导致严重的物理后果。为此,提出一种配电网信息物理系统网络攻击跨信息物理空间传递的风险量化评估方法。针对配电子站,根据IEC 61850标准建立三层信息模型,以断路器和分段开关为攻击目标,分析可能的攻击路径构造攻击图,提出基于相对熵(Kullback-Leibler(KL)距离)并结合逼近理想解排序法(TOPSIS)建立的KL-TOPSIS体系对信息系统脆弱性因子进行量化,结合马尔可夫链和深度优先算法综合计算攻击目标的攻击概率,并建立物理设备遭受网络攻击造成的物理后果指标,分别计算攻击各个断路器和分段开关的物理后果。结合物理后果和攻击概率,得到不同配电终端和配电子站的风险值。最后,以改进的配电网算例验证了所提风险评估模型的有效性。
Abstract:
      Reliable and safe operation of modern physical power system depends on relevant information system. Cyber-attacks against information system can lead to serious physical consequences. So a quantitative risk assessment method of cyber-physical system for cyber-attack in distribution system across cyber and physical spaces is proposed. According to IEC 61850 standard, a three-tier cyber model is established for slave stations in distribution automation system. The possible attack paths are analyzed to construct attack graph, which regards breaker and sectional switches as attack targets. A Kullback-Leibler(KL)-technique for order preference by similarity to an ideal solution(KL-TOPSIS)system based on relative entropy(Kullback-Leibler distance)and TOPSIS is proposed to quantify the vulnerability factors of cyber systems, and Markov chain and depth-first algorithms are combined to synthetically calculate the attack probability of attack targets. Physical consequence indices of physical devices attacked by cyber are established to calculate the physical consequences after attacking breakers and sectional switches. Combining the physical consequences and attack probability, the cyber-attack risk values of different distribution terminals and substations are obtained. Finally, the effectiveness of the proposed risk assessment model is verified by a modified case in distribution system.
查看全文(Free!)   查看附录   查看/发表评论  下载PDF阅读器