基于油液分析的设备状态监测与磨粒识别系统开发

发布时间：2018-06-22 18:58

本文选题：油液分析 + 状态监测　；参考：《华南理工大学》2011年硕士论文

【摘要】：随着国民经济的发展和科学技术水平的不断进步,机械设备朝着大型化、复杂化、连续长周期运行方向发展,因此开展机械设备状态监测和故障诊断显得越来越重要。润滑油是机械设备的血液,它的状态与设备磨损状态最为密切。通过对润滑油的检测分析,可以很好的对设备的运行状态进行监测。本文以油液分析为基础,对某石化公司的大型超高压锅炉给水泵设备进行长期监测,并开发了一套磨粒识别系统软件,该软件可以对几种常见磨粒进行识别,辅助油液监测的展开。在对设备进行监测的过程中,采用了铁谱分析与油液理化指标分析的综合分析方法。一方面利用趋势图三线值法、铁谱磨粒的定性和定量分析描述了设备的磨损状态;另一方面又通过监测油液的粘度、水分、酸值、破乳化值、机械杂质、闪点6项理化指标,分析了油品的质量与衰变程度,确保设备良好的润滑状态。在为期一年的监测中,通过油液分析及时发现问题并解决问题,保证了设备的润滑状态和正常磨损状态,使设备能够良好的运行。在现代油液分析中,主要通过磨粒来判断设备故障程度和失效类型。由于传统的磨粒分析与识别依赖于分析人员的知识和经验,大大限制了油液分析技术的发展。本文以Visual C++ 6.0为软件开发平台,开发了一套磨粒识别系统。这套系统结合了油液分析技术、图像处理技术和智能识别技术,具有界面友好,操作简单,运行效率高,兼容性和扩展能力强等特点。要完成对磨粒的识别,必须解决铁谱磨粒图像中磨粒与背景以及磨粒与磨粒之间的分割问题。通过图像缩放、灰度化、二值化、形态学运算、轮廓提取、链码提取、种子填充等图像处理算法的结合,很好的解决了磨粒图像的分割问题。最后,模仿人对物体的认识,提取了磨粒的形状、纹理以及颜色特征,结合BP神经网络分类算法,成功地对5种典型磨粒(正常磨粒、切削磨粒、球状磨粒、严重滑动磨粒、疲劳磨损磨粒)进行分类识别。
[Abstract]:With the development of national economy and the progress of science and technology, mechanical equipment is developing towards the direction of large scale, complication and continuous long period operation. Therefore, it is more and more important to carry out condition monitoring and fault diagnosis of machinery and equipment. Lubricating oil is the blood of mechanical equipment, its condition and equipment wear state most closely. Through the detection and analysis of the lubricating oil, the running state of the equipment can be monitored well. On the basis of oil analysis, this paper has carried out long-term monitoring of feedwater pump equipment of a large ultra-high pressure boiler in a petrochemical company, and developed a set of abrasive particle recognition system software, which can identify several kinds of common abrasive particles. Development of auxiliary oil monitoring. In the process of monitoring the equipment, a comprehensive analysis method of ferrography analysis and oil physicochemical index analysis was adopted. On the one hand, the wear state of the equipment is described qualitatively and quantitatively by using the three line value method of trend diagram, on the other hand, the physical and chemical indexes of oil viscosity, moisture, acid value, breaking emulsification value, mechanical impurity and flash point are monitored. The quality and decay degree of the oil are analyzed to ensure the good lubrication condition of the equipment. In the course of monitoring for one year, the problems can be found and solved in time through oil analysis, which ensures the lubrication state and normal wear state of the equipment and makes the equipment run well. In modern oil analysis, the fault degree and failure type of equipment are judged by abrasive particles. The traditional analysis and identification of abrasive particles depend on the knowledge and experience of the analysts, which greatly limits the development of oil analysis technology. In this paper, a wear particle recognition system is developed on the platform of Visual C 6.0. The system combines oil analysis technology, image processing technology and intelligent recognition technology. It has the advantages of friendly interface, simple operation, high operation efficiency, strong compatibility and expansibility. In order to recognize the wear particles, we must solve the problem of the segmentation between the wear particles and the background and between the wear particles and the wear particles in the image of ferrographic wear particles. Through the combination of image scaling, grayscale, binarization, morphological operation, contour extraction, chain code extraction, seed filling and other image processing algorithms, the problem of abrasive image segmentation is well solved. Finally, the shape, texture and color features of abrasive particles are extracted by imitating people's understanding of objects. Combined with BP neural network classification algorithm, five typical wear particles (normal wear particles, cutting wear particles, ball wear particles, serious sliding wear particles) are successfully processed. Fatigue wear abrasives) are classified and identified.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3

【参考文献】