基于滤膜阻塞技术和液位图像处理的油液污染度检测方法  

Oil Pollution Detection Based on Filter Blockage Technique and Liquid Level Image Processing

作　　者：白亚洲 郭瑞宇 卢继霞[1] 黄俊凯 程炼 朱成现 BAI Ya-zhou;GUO Rui-yu;LU Ji-xia;HUANG Jun-kai;CHENG Lian;ZHU Cheng-xian(School of Mechanical Electronic and Information Engineering,China University of Mining and Technology(Beijing),Beijing 100083;Ziguangyun Technology Co.,Ltd.,Tianjin 300450)

机构地区：[1]中国矿业大学(北京)机电与信息工程学院,北京100083 [2]紫光云技术有限公司,天津300450

出　　处：《液压与气动》2020年第9期14-21,共8页Chinese Hydraulics & Pneumatics

基　　金：国家自然科学基金(51375481);中央高校基本科研业务费(2013-YJ01)。

摘　　要：为提高液压油污染度检测的智能化,提出基于滤膜阻塞技术和液位图像处理相结合进行油液污染度检测的方法。制定了油液污染度检测的总体方案,对用到的部分元件行了简单介绍。研究了利用图像处理技术获得被测液在恒压下通过滤膜的流量衰减曲线的方法,建立了流量衰减曲线与被测液颗粒浓度关系的数学模型,介绍了利用该模型进行污染度定量检测的方法。对同一被测液分别利用颗粒计数器法和该方法进行污染度检测,测试结果表明,尽管颗粒尺寸分布数据和计数器检测结果存在一定误差,但测得的污染度等级相同,说明该方法及检测模型可以满足工程中对油液污染度检测精度的要求。In order to improve the intelligence of hydraulic oil detection,a method of oil contamination detection based on the combination of the filter blockage technique and liquid level image processing is proposed.An oil pollution detection overall scheme is formulated,and some components used are briefly introduced.The method of using image processing technology to obtain the flow attenuation curve of the measured liquid through the filter membrane under constant pressure is studied.The mathematical model of the relationship between the flow attenuation curve and the concentration of the measured liquid particles is established,and the quantitative detection method for oil sample by using this model is also introduced.The particle counter method and the method mentioned in this article are used to detect the pollution degree of the same oil sample.The test results showed that although the particle size distribution data have certain errors,the pollution degree levels are the same,which indicating that the method and detection model proposed in this paper can meet the requirements for the accuracy of oil pollution detection in engineering.

关 键 词：油液污染度检测 滤膜阻塞技术 图像处理 定量检测 

分 类 号：TH137[机械工程—机械制造及自动化] TH39