Paper Info
Title
Anomaly Detection for Power Grid Flow Patterns Based on the Multi-restricted Boltzmann Machines
Abstract
The power grid is an important facility, which is essential to the national economy and people's livelihood. With the development of the power grid, the security problems of it become more and more urgent. Among these problems, the traffic anomaly of the power grid is one of the most important in the power grid security field currently. It is obvious that the grid data are quite regular, massive, unsupervised and real-time. In addition, the existing methods of this problem are relatively few and the application of those in the specific grid environment is unsatisfactory. Given those factors, this paper proposed a new method of the network anomaly traffic detection, which is based on a benchmark model built by the multi-RBM network. This method is as follows: first, divide the data onto clusters according to the time. Then, determine the RBM model of the initial clusters. Next, obtain the RBM model of the entire data by merging the similar clusters. Finally, detect anomalies according to the similarity between the real-time data and the normal benchmark model. The results of the experiment indicate that the method proposed in this paper is more accurate and responsive in the anomaly detection of power grid flow in comparison with the previous unsupervised algorithms.
Year
DOI
Venue
2018
10.1109/DSC.2018.00100
2018 IEEE Third International Conference on Data Science in Cyberspace (DSC)
Keywords
Field
DocType
Anomaly detection,Restricted Boltzmann Machine,Intrusion detection,clustering
Restricted Boltzmann machine,Data modeling,Data mining,Anomaly detection,Boltzmann machine,Computer science,Cluster analysis,Intrusion detection system,Grid,Benchmark (computing)
Conference
ISBN
Citations 
PageRank 
978-1-5386-4211-5
0
0.34
References 
Authors
3
4
Name
Order
Citations
PageRank
Yichen Li125.79
Yinghua Ma2377.12
Mingda Guo300.34
Shenghong Li435747.31