Title :
Void fraction measurement of gas-liquid two-phase flow in mini-pipe based on image sequence
Author :
Haifeng Ji ; Bei Jiang ; Zhiyao Huang ; Baoliang Wang ; Haiqing Li
Author_Institution :
Dept. of Control Sci. & Eng., Zhejiang Univ., Hangzhou, China
Abstract :
Based on image sequence, a void fraction measurement model of gas-liquid two-phase flow in mini-pipe is developed using support vector regression (SVR) and particle swarm optimization (PSO). A high-speed image acquisition system is constructed to capture dynamic gray image sequence of gas-liquid two-phase flow. The area ratio of gas phase in longitudinal section of the pipe for every image of image sequence is calculated and one-dimension time series can be obtained. And then the mean value, the standard deviation and the nonsymmetrical coefficient are extracted from the one-dimension time series as input vector of the void fraction measurement model. The experiment is carried out in the horizontal mini-pipe with inner diameter of 4.0mm. The results show that the presented void fraction measurement model is feasible and effective. The maximum relative errors of void fraction of slug flow and bubbly flow are less than 8%.
Keywords :
bubbles; computational fluid dynamics; flow visualisation; image sequences; particle swarm optimisation; pipe flow; regression analysis; support vector machines; time series; two-phase flow; PSO; SVR; bubbly flow; dynamic gray image sequence; gas phase area ratio; gas-liquid two-phase flow; high-speed image acquisition system; horizontal minipipe; inner diameter; input vector; maximum relative errors; nonsymmetrical coefficient; one-dimension time series; particle swarm optimization; pipe longitudinal section; size 4.0 mm; slug flow; standard deviation; support vector regression; void fraction measurement model; Fluid flow measurement; Image sequences; Phase measurement; Support vector machines; Testing; Time series analysis; Vectors; gas-liquid two-phase flow; image sequence; mini-pipe; particle swarm optimization; support vector regression; void fraction measurement;
Conference_Titel :
Imaging Systems and Techniques (IST), 2013 IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4673-5790-6
DOI :
10.1109/IST.2013.6729678
