ass日本风韵熟妇pics男人扒开女人屁屁桶到爽|扒开胸露出奶头亲吻视频|邻居少妇的诱惑|人人妻在线播放|日日摸夜夜摸狠狠摸婷婷|制服 丝袜 人妻|激情熟妇中文字幕|看黄色欧美特一级|日本av人妻系列|高潮对白av,丰满岳妇乱熟妇之荡,日本丰满熟妇乱又伦,日韩欧美一区二区三区在线

The automobile braking liquid is in a gas-liquid two-phase state. The gas-liquid two-phase flow in hydraulic braking pipeline is complicated and difficult to be detected. To detect the flow patterns in hydraulic braking pipeline effectively, an experimental bench was built up by mounting a transparent quartz pipe in a hydraulic braking system of automobile, and a flow pattern identification method was proposed based on the gray level co-occurrence matrix and hierarchical cluster analysis. Totally 160 images of gas-liquid two-phase flow patterns in the hydraulic braking pipeline were captured by a digital high speed camera, the feature parameters of gray level co-occurrence matrix of images were extracted by using image processing techniques, and then these feature parameters were used as source data, which were analyzed by the hierarchical cluster analysis method. In order to improve the accuracy of cluster analysis, different sample spaces and class spaces were compared to find the best combination of sample space and class space. Finally, the flow pattern intelligent identification was realized. The test results indicated that the brake fluid was a gas-liquid two-phase flow in hydraulic braking pipeline of automobile, whose flow patterns were recognized as bubbly flow, plug flow, slug flow and annular flow. The cluster analysis results also proved that the method was successful to identify the four typical flow patterns in hydraulic braking pipeline of automobile, and the whole identification accuracy was up to 95.625%. Especially for bubble flow and slug flow, their identification accuracies were 100%. Due to its high speed and high accuracy, the method provided an effective way for researching the flow patterns of gas-liquid two-phase flow in the hydraulic braking system.