Concentration Measurement of Two-Phase Flow by Means of LS-SVM

Author

Peng, Zhen-Rui ; Liu, Shi-zhong

Author_Institution

Sch. of Mechatronics Eng., Lanzhou Jiaotong Univ.

fYear

2006

fDate

13-16 Aug. 2006

Firstpage

3667

Lastpage

3670

Abstract

Combining least squares support vector machine (LS-SVM) and the 66 capacitance values from capacitance sensor of electrical capacitance tomography (ECT), a new method for phase concentration measurement of two-phase flow was proposed. LS-SVM was used to establish the model for the phase concentration measurement of two-phase flow. In this model, the 66 independent capacitances were the inputs and the concentration was the output. In the measurement process, the capacitance values were obtained by ECT capacitance sensor; then the normalized capacitance values were fed to the model and the concentration value was calculated. This new approach implemented direct concentration measurement of two-phase flow without the need of conventionally time-consuming image reconstruction. The experimental results show the effectiveness of our proposed method

Keywords

capacitance; capacitive sensors; chemical variables measurement; computational fluid dynamics; flow measurement; least squares approximations; phase measurement; support vector machines; tomography; two-phase flow; ECT capacitance sensor; LS-SVM; electrical capacitance tomography; least squares support vector machine; two-phase flow phase concentration measurement; Capacitance measurement; Capacitive sensors; Dielectric measurements; Electrical capacitance tomography; Fluid flow measurement; Image reconstruction; Least squares methods; Permittivity measurement; Phase measurement; Sensor systems; Support vector machines; Concentration; Electrical capacitance; Least squares support vector machine; Two-phase flow;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2006 International Conference on

Conference_Location

Dalian, China

Print_ISBN

1-4244-0061-9

Type

conf

DOI

10.1109/ICMLC.2006.258623

Filename

4028707