| 引用本文: | 杨杉,谭博,郭静波.基于双马尔科夫链的新型能源互联网虚假数据注入攻击检测[J].电力自动化设备,2021,41(2): |
| YANG Shan,TAN Bo,GUO Jingbo.Detection of false data injection attack for new-type energy internet based on double Markov chains[J].Electric Power Automation Equipment,2021,41(2): |
|
| 摘要: |
| 新型能源互联网中各环节信息的高度融合和开放削弱了其防御外界攻击的能力。针对具有隐蔽性特征的虚假数据注入攻击(FDIA)，提出了适用于新型能源互联网的基于双马尔科夫链的FDIA检测方法。考虑到新型能源互联网的FDIA原理、特征以及新型能源互联网包含大量的测量数据和多变的运行状态，将待检测的数据映射到2个不同的状态空间，并生成2个不同的马尔科夫链模型；根据该模型估计所得的能源互联网运行状态的精确度，生成FDIA的检测器。通过实验案例验证所提检测方法的正确性和有效性。实验结果表明：所提FDIA检测方法具有优秀的检测概率和较少的检测计算量，检测概率可达98.60%，虚警概率仅为1.35 %，计算量相比于支持向量机方法降低了1个数量级，能满足新型能源互联网的应用要求。 |
| 关键词: 新型能源互联网 信息安全 攻击检测 虚假数据注入攻击 马尔科夫链 |
| DOI：10.16081/j.epae.202101001 |
| 分类号:TM761 |
| 基金项目:国家自然科学基金资助项目(51677094)；国家电网有限公司总部科技项目(面向5G的信息能源融合技术与装备研制) |
|
| Detection of false data injection attack for new-type energy internet based on double Markov chains |
|
YANG Shan, TAN Bo, GUO Jingbo
|
|
Department of Electrical Engineering, Tsinghua University, Beijing 100084, China
|
| Abstract: |
| The highly integrated and open information in each link of new-type energy internet weakens its ability to defend against external attacks. Aiming at the FDIA(False Data Injection Attack) with the characteristics of concealability, a FDIA detection method available for new-type energy internet is proposed based on double Markov chains. Considering the principle and characteristics of FDIA and a large number of measurement data and changeable operation states in the new-type energy internet, the data to be detected is mapped into two different state spaces and two different Markov chain models are generated. Based on the accuracy of energy internet’s operation state estimated by the model, FDIA detector is generated. The correctness and effectiveness of the proposed detection method are verified by experimental cases. The experimental results show that the proposed FDIA detection method has excellent detection probability and less detection computation amount, the detection probability is up to 98.60%,and the false alarm probability is only 1.35 %. The calculation amount is reduced by one order of magnitude compared with the support vector machine method, which can meet the application requirements of new-type energy internet. |
| Key words: new-type energy internet information security attack detection false data injection attack Markov chain |
