基于电容层析成像的气液两相流流型聚类分析

• 作 者: 杜运成 王化祥
• 机 构: 天津大学电气与自动化工程学院,
• 出 处: 中国电机工程学报
• 发表日期: 2009年29期
• 关键词: electrical capacitance tomography flow regime identification wavelet package ene

气液两相流流型对截面含气率等流动参数测量影响很大。基于电容层析成像(electrical capacitance tomography,ECT)系统,以水平管气液两相流流型为研究对象,利用测量电容信号包含流型信息的特点,采用小波包能量信息熵直接提取不同流型的特征参数,进而结合模糊c均值聚类(fuzzyc-means,FCM)算法实现流型识别。实验结果表明,对于水平管层流、满管流、塞状流和弹状流典型气液两相流流型,该方法具有较高的识别率。

The flow pattern of gas-liquid two-phase flow has a great influence on the measurement of flow parameters such as gas-bearing cross-section. Based on the capacitive tomography (ECT) system, taking the flow pattern of gas-liquid two-phase flow in a horizontal pipe as the research object, this paper uses the characteristic that the measured capacitance signal contains the flow pattern information, and uses the wavelet packet energy information entropy to directly extract different flow patterns And then combined with fuzzy c-means (FCM) algorithm to realize the flow pattern recognition. The experimental results show that this method has high recognition rate for the typical gas-liquid two-phase flow pattern of horizontal tube laminar flow, full tube flow, slug flow and slug flow.