An Online Flow Pattern Identification System for Gas–Oil Two-Phase Flow Using Electrical Capacitance Tomography

摘要

An online and noninvasive gas–oil two-phase flow pattern identification system is proposed based on the electrical capacitance tomography (ECT) technique. A fuzzy pattern recognition technique is introduced to solve the flow pattern identification problem because the flow pattern and its transition are fuzzy in nature. Images captured by the ECT system are analyzed through image processing and a pattern recognition technique. The flow pattern identification results are achieved by fuzzy sets theory. The online flow pattern identification system is established, and typical flow patterns, such as the homogeneous flow, the stratified flow, the annular flow, and the slug flow, in horizontal pipeline can be identified. The experimental results show that the online flow pattern identification system is effective.