hybrid model is established for leak detection of longdistance gas transport pipelines. Firstly, a mechanism model is built on the basic transportflow equations, where the mass balance condition, momentum balance condition and state equation are considered. Next, a neural network model is used to compensate for the error of the mechanism model and improve the modelling precision. Here, the radial basis function (RBF) neural network is adopted. Therefore, the merits of the mechanism model and the neural network model are integrated to construct a hybrid model of long-distance gas pipelines. The experimental system of a long-distance pipeline is established and the pressure data of multiple nodes is collected. Finally, based on the experimental pressure data, the output of the mechanism model and the output of the hybrid model are compared. The comparison shows that the detection precision of the hybrid model is better.
Hybrid modelling for leak detection of long-distance gas transport pipeline
Insight ; 55 , 7 ; 372-381
2013
10 Seiten, 30 Bilder, 4 Tabellen, 36 Quellen
Article (Journal)
English
Post-earthquake pipeline leak detection technologies
Tema Archive | 2008
|Study outlines pipeline leak-detection selection
Tema Archive | 2012
|Canadian pipeline installs leak-detection system
Tema Archive | 1988
|Model based pipeline monitoring with leak detection
Tema Archive | 2009
|