A Multilevel Deep Learning Method for Data Fusion and Anomaly Detection of Power Big Data

DOI

10.2991/eeeis-17.2017.79How to use a DOI?

Keywords

power big data, restricted Boltzmann machine, recurrent neural networks, anomaly detection, deep learning, data fusion.

Abstract

With the expansion of the power information network scale, various network threats are also increasing. In order to excavate security threats in power grid by making full use of heterogeneous data sources in power big data, this paper maps heterogeneous data in different formats to a unified embedded vector space with deep restricted Boltzmann machine, and achieves the fusion of heterogeneous data sources. Then, it draws a profile for embedded vector dataset using recurrent neural networks, and achieves the anomaly detection of big data. Experimental results show that the proposed anomaly detection approach has the biggest value in our proposed mutual information metric, and it is obviously better than other anomaly detection algorithms in accuracy, false positive rate and false negative rate. The method of this paper can effectively detect the security threat in the power grid, and it is conducive to the safe and stable operation of power grids.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Dong-Lan LIU
AU  - Xin LIU
AU  - Hao YU
AU  - Wen-Ting WANG
AU  - Xiao-Hong ZHAO
AU  - Jian-Fei CHEN
PY  - 2017/09
DA  - 2017/09
TI  - A Multilevel Deep Learning Method for Data Fusion and Anomaly Detection of Power Big Data
BT  - Proceedings of the 3rd Annual International Conference on Electronics, Electrical Engineering and Information Science (EEEIS 2017)
PB  - Atlantis Press
SP  - 533
EP  - 539
SN  - 2352-5401
UR  - https://doi.org/10.2991/eeeis-17.2017.79
DO  - 10.2991/eeeis-17.2017.79
ID  - LIU2017/09
ER  -