The Gas User Monitoring System
Zhu Lian;Zhao Yong;Zhang Haijun;Jiang Zhong Yu;Wu Jun;College of Electronics and Information Engineering,Sichuan University;Chengdu Qianjia Technology Co.,Ltd;Sichuan Huayou Group Co.,Ltd;
This paper presents a system architecture and corresponding algorithm for monitoring gas consumption by gas users. Based on the relational database commonly used by gas companies, the system converts real-time business data into a distributed system represented by Hadoop ecosphere through data extraction, data cleaning and data conversion(ETL). The system is based on Spark self-built classification algorithm and prediction algorithm to complete the user-side gas monitoring, and return the abnormal results to the client. The monitoring results of industrial and commercial users in Chengdu City Gas Company and CNGC show that the system has the advantages of high efficiency and accuracy.
【CateGory Index】： TU996.7