Applying Deep Learning Method to Data Analysis of Low Voltage Line Carrier Module

2018 5th International Conference on Advanced Composite Materials and Manufacturing Engineering

Cao, Guorui^1 ; Yang, Li^1 ; Yu, Xuejun^1 ; Zhong, Ruijun^1 ; Lu, Jinyu^1

Electric Power Science and Research Institute of Tianjin Electric Power Company, Tianjin, China^1

关键词: Business Process;    Detection technology;    Learning methods;    Learning network;    Low-voltage line;    Low-voltage power line carrier communications;    Smart energy meters;    Technical solutions;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/394/4/042095/pdf
DOI  :  10.1088/1757-899X/394/4/042095

来源: IOP

【 摘 要 】

With the increasing popularity of the global energy Internet, the smart grid in the city has developed rapidly, and the number of smart energy meters is increasing year by year. The demand for detecting the carrier module of the power meter has also increased to a new level. But the existing carrier module data analysis scheme is not perfect. It is urgent to need a complete set of carrier module data analysis scheme and new detection technology to realize the multiplexing of carrier module data, and to excavate the value of the carrier module to detect the data. Unlike previous studies, which using SVM and Bayesian networks, this paper proposes a scheme for classification of carrier module detection data using a deep learning network. GRU deep neural network is used to model the test data and identify the environment automatically. According to the recognition results, the carrier module can be adjusted accordingly. Experiments show that the scheme proposed in this paper has good results. Our scheme has very good effect in low voltage power line carrier communication data analysis task. It can excavate valuable information of low voltage power line carrier communication and has high economic value. In addition, the technical solutions proposed in this paper will not affect the existing business processes, and can be applied in large areas.