Title :
A method of two-phase flow recognition based on dynamic cluster in horizontal pipe
Author :
Hua, Shuang ; Zhao, Fu-Lai ; Xu, Yan-Bin ; Dong, Feng
Author_Institution :
Sch. of Electr. & Autom. Eng., Tianjin Univ., China
Abstract :
The accurate measurement of the parameters of two-phase flows depends on flow regime recognition. Based on the application of electrical resistance tomography (ERT) and the flow regimes´ diversity in horizontal pipe, variance ratio (C.V.) as an eigenvalue of cross section has been used to express the information of gas-liquid two-phase flow in measured field. Furthermore, the difference of the C.V. eigenvalues of cross sections provides an evidence of dynamic clustering to recognize several typical flow regimes. It is feasible to take the method to recognize flow regimes by validation of the practical experiment.
Keywords :
computerised tomography; eigenvalues and eigenfunctions; electric resistance measurement; pattern clustering; two-phase flow; dynamic cluster; dynamic clustering; eigenvalue; electrical resistance tomography; flow regime diversity; flow regime recognition; gas-liquid flow; horizontal pipe; two-phase flow recognition; Automation; Eigenvalues and eigenfunctions; Electric resistance; Electric variables measurement; Electrical resistance measurement; Fluid dynamics; Fluid flow measurement; Petroleum; Power engineering and energy; Tomography; ERT; Two-phase flow; dynamic cluster; flow regime recognition; variance ratio;
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
DOI :
10.1109/ICMLC.2005.1527231
