Intelligent Pipeline Leak Detection and Analysis System
Hui‐Wen Lin, Hezhi Lin, Xiaoxuan Fang, Mingkang Wang, Lianfen Huang
Abstract
Urban water supply pipe network leakage accident is a joint problem that exists among the water supply industries all over the world. The leakage in the pipe network not only causes waste of water resources and energy, but also brings water supply security risks to the entire pipe network. Therefore, monitoring the leakage in the water supply system is an important component of water work nowadays. This paper presents an intelligent pipeline leak detection and analysis system. The proposed system collects pipe flow data with an ultrasonic flowmeter. This system uses NB-IoT for data transmission and One-Class-SVM's outlier detection method for leakage judgment. It also gives a rough position of leakage. Compared with the traditional leakage detection method, the proposed system has many advantages such as simple equipment, convenient operation and access without breaking the original pipe section. Besides, the proposed system provides a friendly web interface and Android App interface, allowing users to observe real-time pipe flow data. It also has an alarm so that users will know whether there has leakage.