基于多视角低秩分析的电力状态不良数据检测
|
摘要
随着信息化技术在智能电网的应用逐步深入,在智能电网的运维中能及时自动检测到不良数据,如网络攻击数据和设备故障数据,对电网的稳定和持续运行有着重要意义。该文提出一种基于多视角低秩分析的电力状态不良数据检测算法。该算法使用来自多个观测源的观测数据综合估计电力系统的状态,算法使用低秩模型挖掘出来自多个观测源数据间的共享本真数据,同时使用稀疏模型对不良数据建模。针对所提出的目标方程,给出了一种基于交叉迭代的优化算法。最后,在IEEE多个节点测试系统上的实验证明了该算法相对于已有算法的先进性。
With the widely deployment of information techniques in smart grid, it is quite important to automatically detect the bad data, e.g., malicious injection data and unfunctional sensor data, from daily observations. In this paper, we propose a novel approach for bad data detection in smart grid based on multi-view low-rank analysis. Specifically, the proposed method estimates the grid state by analyzing the data collected from multiple sources. A low-rank function is learned to unveil the shared true data from observations, and the sparsity of data is applied to formulate bad data. Furthermore, an iterative optimization algorithm is proposed to solve the objective function. At last, extensive experiments on several IEEE bus systems verify the superiority of the proposed method.
With the widely deployment of information techniques in smart grid, it is quite important to automatically detect the bad data, e.g., malicious injection data and unfunctional sensor data, from daily observations. In this paper, we propose a novel approach for bad data detection in smart grid based on multi-view low-rank analysis. Specifically, the proposed method estimates the grid state by analyzing the data collected from multiple sources. A low-rank function is learned to unveil the shared true data from observations, and the sparsity of data is applied to formulate bad data. Furthermore, an iterative optimization algorithm is proposed to solve the objective function. At last, extensive experiments on several IEEE bus systems verify the superiority of the proposed method.
引文
[1]陈树勇,宋书芳,李兰欣,等.智能电网技术综述[J].电网技术,2009(8):1-7.CHEN Shu-yong,SONG Shu-fang,LI Lan-xin,et al.Asurvey on smart grid[J].Power System Technology,2009,33(8):1-7.
[2]CUI Shu-guang,HAN Zhu,KAR S,et al.Coordinated data-injection attack and detection in the smart grid:a detailed look at enriching detection solutions[J].IEEESignal Processing Magazine,2012,29(5):106-115.
[3]HUANG Yi,MOHAMMAD E,NGUYEN H,et al.Bad data injection in smart grid:attack and defense mechanisms[J].IEEE Communications Magazine,2013,51(1):27-33.
[4]MOHAMMAD E,HAN Zhu,SONG Ling-yang.Effect of stealthy bad data injection on network congestion in market based power system[C]//Wireless Communications and Networking Conference.Shanghai,China:IEEE,2012:2468-2472.
[5]LIU Guang-can,LIN Zhou-chen,YAN Shui-cheng,et al.Robust recovery of subspace structures by low-rank representation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(1):171-184.
[6]LI Jing-jing,WU Yue,ZHAO Ji-dong,et al.Low-rank discriminant embedding for multiview learning[J].IEEETransactions on Cybernetics,2017,47(11):3516-3529.
[7]LIU Lan-chao,ESMALIFALAK M,HAN Zhu.Detection offalse data injection in power grid exploiting low rank and sparsity[C]//2013 IEEE International Conference on Communications.Budapest,Hungary:IEEE,2013:4461-4465.
[8]JOHN W,GANESH A,RAO S,et al.Robust principal component analysis:exact recovery of corrupted low-rank matrices via convex optimization[R].Urbana-Champaign:Coordinated Science Laboratory,University of Illinois,2009.
[9]LIN Zhou-chen,CHEN Min-ming,MA Yi.The augmented lagrange multiplier method for exact recovery of corrupted low-rank matrices[R].Urbana-Champaign:Coordinated Science Laboratory,University of Illinois,2009.
[10]BOYD S,PARIKH N,CHU E,et al.Distributed optimization and statistical learning via the alternating direction method of multipliers[J].Foundations and Trends in Machine Learning,2011,3(1):1-122.
[11]CAI Jian-feng,CANDES E,SHEN Zuo-wei.A singular value thresholding algorithm for matrix completion[J].SIAM Journal on Optimization,2010,20(4):1956-1982.
[12]OLIVER K,JIA Li-yan,THOMAS R,et al.Malicious data attacks on the smart grid[J].IEEE Transaction on Smart Grid,2011,2(4):645-658.
[13]ZHAO Han-dong,FU Yun.Dual-regularized multi-view outlier detection[C]//International Conference on Artificial Intelligence.Buenos Aires,Argentina:AAAI Press,2015:4077-4083.
[2]CUI Shu-guang,HAN Zhu,KAR S,et al.Coordinated data-injection attack and detection in the smart grid:a detailed look at enriching detection solutions[J].IEEESignal Processing Magazine,2012,29(5):106-115.
[3]HUANG Yi,MOHAMMAD E,NGUYEN H,et al.Bad data injection in smart grid:attack and defense mechanisms[J].IEEE Communications Magazine,2013,51(1):27-33.
[4]MOHAMMAD E,HAN Zhu,SONG Ling-yang.Effect of stealthy bad data injection on network congestion in market based power system[C]//Wireless Communications and Networking Conference.Shanghai,China:IEEE,2012:2468-2472.
[5]LIU Guang-can,LIN Zhou-chen,YAN Shui-cheng,et al.Robust recovery of subspace structures by low-rank representation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(1):171-184.
[6]LI Jing-jing,WU Yue,ZHAO Ji-dong,et al.Low-rank discriminant embedding for multiview learning[J].IEEETransactions on Cybernetics,2017,47(11):3516-3529.
[7]LIU Lan-chao,ESMALIFALAK M,HAN Zhu.Detection offalse data injection in power grid exploiting low rank and sparsity[C]//2013 IEEE International Conference on Communications.Budapest,Hungary:IEEE,2013:4461-4465.
[8]JOHN W,GANESH A,RAO S,et al.Robust principal component analysis:exact recovery of corrupted low-rank matrices via convex optimization[R].Urbana-Champaign:Coordinated Science Laboratory,University of Illinois,2009.
[9]LIN Zhou-chen,CHEN Min-ming,MA Yi.The augmented lagrange multiplier method for exact recovery of corrupted low-rank matrices[R].Urbana-Champaign:Coordinated Science Laboratory,University of Illinois,2009.
[10]BOYD S,PARIKH N,CHU E,et al.Distributed optimization and statistical learning via the alternating direction method of multipliers[J].Foundations and Trends in Machine Learning,2011,3(1):1-122.
[11]CAI Jian-feng,CANDES E,SHEN Zuo-wei.A singular value thresholding algorithm for matrix completion[J].SIAM Journal on Optimization,2010,20(4):1956-1982.
[12]OLIVER K,JIA Li-yan,THOMAS R,et al.Malicious data attacks on the smart grid[J].IEEE Transaction on Smart Grid,2011,2(4):645-658.
[13]ZHAO Han-dong,FU Yun.Dual-regularized multi-view outlier detection[C]//International Conference on Artificial Intelligence.Buenos Aires,Argentina:AAAI Press,2015:4077-4083.