基于高速摄影传感器的气液两相流型分层模糊识别

摘要

Objective and intelligent identification of gas-liquid two phase flow pattern have important meaning for other parameters measurement. High-speed camera is used as detecting sensor to capture two-phase flow patterns image in adiabatic vertical section whose inner diameter is 50mm,extract five eigenvectors of mean and standard deviation of gray image,and area,width,height of the max object of binary image;use hierarchical fuzzy reasoning method to identify bubble flow,slug flow,plug flow and churn flow patterns. For 131 experimental images,the average identification accuracy of this method is 93. 13 % .%气液两相流流型的客观和智能识别对两相流其它参数测量具有重要意义。用高速摄像机作为检测传感器，获取内径为50 mm 的绝热垂直管段内气液两相流流型图像，提取流型灰度图像的均值、标准差及二值图像的最大物体面积、宽度和高度五个特征向量，采用分层模糊推理方法实现泡状流、块状流、塞状流和乳沫状流的流型识别。对于131幅实验图像，该方法的识别正确率平均达到93.13%。