一种基于小生境遗传算法的电网信息安全风险评估模型

doi:10.12204/j.issn.1000-7229.2021.03.011

摘要/Abstract

摘要：

信息安全保障对于电力信息物理系统安全稳定运行至关重要,其关键在于对电力信息物理系统进行全方位实时监控,并对采集到的海量监测数据进行分析以做出准确的安全风险评估结果。作为用于模式分类的进化算法,基因表达式编程(gene expression programming,GEP)算法由于其可以执行全局搜索而受到广泛关注,但其在高维度数据集下的运算极为耗时。针对上述问题,提出了一种基于小生境提高样本多样性的改进基因表达式编程算法用于电网信息安全风险评估,该算法首先利用粗糙集的思想,通过分辨函数求解最优属性对数据样本进行约简,再利用小生境模型提高约简样本个体的多样性以加快GEP算法运算的收敛速度,进而通过遗传算法实现全局搜索并得到安全风险等级评估结果。仿真实验表明,与传统的安全风险评估算法相比,提出的改进GEP算法具有较高的属性约简率和全局收敛率,可以快速实现海量监测数据下的电网信息安全风险评估。

关键词: 电网信息系统, 安全风险评估, 基因表达式编程, 小生境, 属性约简

Abstract:

Information security assurance is essential for the safe and stable operation of power cyber-physical systems. The key is to conduct all-round real-time monitoring of power cyber-physical systems and analyze the collected massive monitoring data to make accurate security risk assessments. As an evolutionary algorithm for pattern classification, the gene expression programming (GEP) has received widespread attention due to its ability to perform global search, but its operation on high-dimensional data sets is extremely time-consuming. This paper proposes an enhanced GEP algorithm for power grid information security risk assessment. The algorithm first uses the idea of rough set, solves the optimal attribute through the discrimination function to reduce the data sample, and then uses the niche genetic algorithm improves the diversity of the reduced sample individuals to accelerate the convergence speed of the GEP algorithm, and then realizes the global search through the genetic algorithm and obtains the level assessment of the security risk. Simulation results show that, compared with traditional security risk assessment algorithms, the enhanced GEP algorithm proposed in this paper has a higher attribute reduction rate and global convergence rate, and can quickly implement risk assessment from massive monitoring data of grid security information.

Key words: power grid information system, security risk assessment, gene expression programming, niche, attribute reduction

中图分类号:

TM769

李佳玮, 吴克河, 张波. 一种基于小生境遗传算法的电网信息安全风险评估模型[J]. 电力建设, 2021, 42(3): 89-96.

LI Jiawei, WU Kehe, ZHANG Bo. A Risk Assessment Model Based on Niche Genetic Algorithm for Power System Information Security[J]. ELECTRIC POWER CONSTRUCTION, 2021, 42(3): 89-96.

图/表 10

图1 安全风险评估模型算法流程图

Fig.1 Algorithm flowchart of security risk assessment model

图2 小生境算法适应度损失率Floss和参数R, V关系

Fig.2 Relationship between niche algorithm fitness loss rate and parameter selection R, V

图3 原始GEP算法和本文算法的适应度损失率比较

Fig.3 Comparison of fitness loss rate between original GEP algorithm and proposed algorithm

图4 AR-DF、PCA以及SVD算法的属性约简率比较

Fig.4 Comparison of attribute reduction rate among AR-DF, PCA and SVD

图5 AR-DF、PCA以及SVD算法在求解最优属性约简的耗时比较

Fig.5 Time-consumption comparison among AR-DF, PCA and SVD

图6 属性约简前后本文算法和传统算法的全局收敛率比较

Fig.6 Comparison of convergence rates between proposed and traditional algorithm and T_GEP

图7 属性约简前后本文算法的风险评估耗时比较

Fig.7 Comparison of risk assessment time-consumption of proposed algorithm before and after attribute reduction

图8 针对训练数据集的基于本文算法的风险评估真实值与模型值比较

Fig.8 Comparison of the real value of risk assessment and model value based on proposed algorithm fortraining data set

图9 针对测试数据集的基于本文算法的风险评估真实值与模型值比较

Fig.9 Comparison between real value and model value of risk assessment based on proposed algorithm for test data set

图10 本文算法和传统遗传算法评估结果比较

Fig.10 Comparison of the evaluation results among three genetic algorithms

参考文献 20

[1]	李培恺, 刘云, 辛焕海, 等. 分布式协同控制模式下配电网信息物理系统脆弱性评估[J]. 电力系统自动化, 2018,42(10):22-29,59.
	LI Peikai, LIU Yun, XIN Huanhai, et al. Vulnerability assessment for cyber physical system of distribution network in distributed cooperative control mode[J]. Automation of Electric Power Systems, 2018,42(10):22-29,59.
[2]	李存斌, 李小鹏, 田世明, 等. 能源互联网电力信息深度融合风险传递:挑战与展望[J]. 电力系统自动化, 2017,41(11):17-25.
	LI Cunbing, LI Xiaopeng, TIAN Shiming, et al. Challenges and prospects of risk transmission in deep fusion of electric power and information for energy internet[J]. Automation of Electric Power Systems, 2017,41(11):17-25.
[3]	张殷, 肖先勇, 李长松. 基于攻击者视角的电力信息物理融合系统脆弱性分析[J]. 电力自动化设备, 2018,38(10):81-88.
	ZHANG Yin, XIAO Xianyong, LI Changsong. Vulnerability analysis of cyber physical power system from attacker’s perspective[J]. Electric Power Automation Equipment, 2018,38(10):81-88.
[4]	蔡星浦, 王琦, 邰伟, 等. 基于多阶段博弈的电力CPS虚假数据注入攻击防御方法[J]. 电力建设, 2019,40(5):48-54.
	CAI Xingpu, WANG Qi, TAI Wei, et al. A multi-stage game based defense method against false data injection attack on cyber physical power system[J]. Electric Power Construction, 2019,40(5):48-54.
[5]	WANG Q, PIPATTANASOMPORN M, KUZLU M, et al. Framework for vulnerability assessment of communication systems for electric power grids[J]. IET Generation, Transmission & Distribution, 2016,10(2):477-486.
[6]	夏卓群, 曾悠优, 尹波, 等. 电力网络中基于物理信息的虚假数据入侵检测方法[J]. 信息网络安全, 2019(4):29-36.
	XIA Zhuoqun, ZENG Youyou, YIN Bo, et al. False data intrusion detection method based on physical information in power network[J]. Netinfo Security, 2019(4):29-36.
[7]	WANG Y N, LIN Z Y, LIANG X, et al. On modeling of electrical cyber-physical systems considering cyber security[J]. Frontiers of Information Technology & Electronic Engineering, 2016,17(5):465-478.
[8]	刘文霞, 宫琦, 郭经, 等. 基于混合通信网的主动配电信息物理系统可靠性评价[J]. 中国电机工程学报, 2018,38(6):1706-1718,1907.
	LIU Wenxia, GONG Qi, GUO Jing, et al. Reliability simulation of ADN cyber-physical system based on hybrid communication network[J]. Proceedings of the CSEE, 2018,38(6):1706-1718,1907.
[9]	YAZDI M, KABIR S. Fuzzy evidence theory and Bayesian networks for process systems risk analysis[J]. Human and Ecological Risk Assessment: An International Journal, 2020,26(1):57-86. doi: 10.1080/10807039.2018.1493679 URL
[10]	殷加玞, 赵冬梅. 基于全概率风险度量的电力系统备用风险评估方法[J]. 电力自动化设备, 2020,40(1):156-162.
	YIN Jiafu, ZHAO Dongmei. Reserve risk assessment method of power system based on total probability risk measure[J]. Electric Power Automation Equipment, 2020,40(1):156-162.
[11]	赵刚, 吴天水. 结合灰色网络威胁分析的信息安全风险评估[J]. 清华大学学报(自然科学版), 2013,53(12):1761-1767.
	ZHAO Gang, WU Tianshui. Information security risk assessment based on G-ANP[J]. Journal of Tsinghua University (Science and Technology), 2013,53(12):1761-1767.
[12]	ISLAM M S, NEPAL M P, SKITMORE M, et al. A knowledge-based expert system to assess power plant project cost overrun risks[J]. Expert Systems With Applications, 2019,136:12-32. doi: 10.1016/j.eswa.2019.06.030 URL
[13]	盛魁, 王伟, 卞显福, 等. 混合数据的邻域区分度增量式属性约简算法[J]. 电子学报, 2020,48(4):682-696.
	SHENG Kui, WANG Wei, BIAN Xianfu, et al. Neighborhood discernibility degree incremental attribute reduction algorithm for mixed data[J]. Acta Electronica Sinica, 2020,48(4):682-696.
[14]	HAJIHASSANI M, ABDULLAH S S, ASTERIS P G, et al. A gene expression programming model for predicting tunnel convergence[J]. Applied Sciences, 2019,9(21):4650 doi: 10.3390/app9214650 URL
[15]	SHUVRO R A, WANGT Z, DAS P, et al. Modeling cascading-failures in power grids including communication and human operator impacts[C]// 2017 IEEE Green Energy and Smart Systems Conference (IGESSC). Long Beach, CA, USA: IEEE, 2017: 1-6.
[16]	DENG S, YUE D, FU X, et al. Security risk assessment of cyber physical power system based on rough set and gene expression programming[J]. IEEE/CAA Journal of Automatica Sinica, 2015,2(4):431-439. doi: 10.1109/JAS.2015.7296538 URL
[17]	LIU F C, HUO W J, HAN Y, et al. Study on network security based on PCA and BP neural network under green communication[J]. IEEE Access, 2020,8:53733-53749. doi: 10.1109/Access.6287639 URL
[18]	TEOH T T, ZHANG Y, NGUWI Y Y, et al. Analyst intuition inspired high velocity big data analysis using PCA ranked fuzzy k-means clustering with multi-layer perceptron (MLP) to obviate cyber security risk[C]// 2017 13th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD). Guilin, China: IEEE, 2017: 1790-1793.
[19]	MALOSSINI A, BLANZIERI E, CALARCO T. Quantum genetic optimization[J]. IEEE Transactions on Evolutionary Computation, 2008,12(2):231-241. doi: 10.1109/TEVC.2007.905006 URL
[20]	BOUKTIF S, FIAZ A, OUNI A, et al. Optimal deep learning LSTM model for electric load forecasting using feature selection and genetic algorithm: Comparison with machine learning approaches[J]. Energies, 2018,11(7):1636-1647. doi: 10.3390/en11071636 URL