Open AccessA Power Metering Pipeline Fault Warning Method Based on Deep Learning
Author(s) -
Shen Zhao,
Yiling Xing,
Xiaohui Chen,
Yongjian Ou,
Wanting Zhang,
Bingfeng Lin
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/740/1/012111
Subject(s) - pipeline (software) , fault (geology) , warning system , artificial neural network , computer science , real time computing , electric power system , data mining , artificial intelligence , power (physics) , engineering , reliability engineering , telecommunications , physics , quantum mechanics , seismology , programming language , geology
With the continuous deepening of the informationization and intelligentization of the electric power industry in China, the fault early warning and diagnosis system of the electric power verification system in China reflects the current situation of insufficient intelligentization. Traditional fault diagnosis system collects operation data of verification system through sensor network, acquisition network and log message technology, and then carries out manual or semi-manual fault detection and processing. There is a small amount of data collected, and the traditional data mining methods such as expert judgment method, decision tree method and SDG model have low efficiency and poor diagnosis effect. In view of this situation, this paper introduces deep learning technology into the verification system of automated pipeline, realizes the integration of fault early warning and diagnosis, builds a fault classification model for automated pipeline based on deep learning neural network, and tests and verifies the effect of model early warning with actual system operation data. The validation of the algorithm gives the result of fault early warning in the form of probability, fully considers the factors affecting the interaction between the pipeline and equipment, has a better effect of fault early warning provides more accurate reference for fault detection and prevention of automated pipeline.