IDENTIFICATION OF GAS-LIQUID TWO-PHASE FLOW REGIMES USING VAGUE SETS AND IMAGE PROCESSING

摘要

A new method was developed for identifying gas-liquid two-phase flow regimes. The method was based on vague sets and image processing techniques. The gas-liquid two-phase flow images were captured by a camera in a transparent horizontal pipe. The characteristics of flow images such as area, width, length, and center coordinates were extracted using image processing techniques. The vague sets models of bubbly, stratified, slug, and annular flow regimes were built according to the typical flow image features. The cosine similarity measure was used to determine the degree of similarity between two vague sets. The flow pattern can be identified accurately according to the maximum cosine similarity degree of vague sets between the test flow regime and the typical flow regimes. The experimental results show that the proposed method is effective to identify the basic flow regimes in a horizontal pipe.