自适应数学形态学在轴承故障诊断中的应用研究

发布时间：2018-03-21 16:47

本文选题：轴承故障诊断　切入点：数学形态学　出处：《武汉科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：轴承是机械设备中的重要零部件，在机械系统中应用广泛，但轴承失效时往往会影响机械运作，严重时可能会引发机械系统瘫痪，甚至发生伤亡事故给企业带来巨大经济损失。根据统计数据显示，因轴承失效而引发的机械故障在总体机械故障中占有较大比重。因此，对滚动轴承进行早期故障监测，找出故障发生的位置，预测故障发展方向，具有十分重要的意义。轴承故障诊断中的重点：从待处理信号中提取轴承故障特征信息，判断轴承故障类型和故障部位。本文在学习了轴承故障的类型和特征的基础上，针对传统形态学在滚动轴承故障信号处理中，结构元素选取困难，降噪效果不理想等问题，提出了广义差值滤波算法和自适应张量形态学算法。本文的主要研究内容： 1、提出了广义形态差值滤波的轴承故障特征提取算法。根据待处理信号的局部特征信息，选取最优结构元素尺度，构建广义差值形态滤波器，可以更好的提取故障特征信息。通过仿真信号和轴承模拟故障实验信号证明了该方法是有效的。 2、在广义形态差值滤波算法的基础上，为了进一步提高形态学的降噪和特征提取效果，提出了一种自适应张量形态学的轴承故障特征提取算法。该算法根据待处理信号的局部特征信息，构建张量椭圆结构元素，能取代传统的直线型结构元素和圆盘结构元素。利用张量形态学滤波器对轴承故障信号进行降噪和提取故障特征，在轴承故障中得到了很好的应用效果。 3、对比三种形态处理方法在轴承故障特征提取中的优劣，综合分析结果表明：自适应张量形态学在轴承内、外圈故障的特征提取中效果最优，其次是广义差值形态学提取效果，最末的是传统形态学提取效果；而针对轴承滚动体故障信号分析，传统形态学提取效果最优，，其次是自适应张量形态学和广义差值形态学。
[Abstract]:Bearing is an important part in mechanical equipment, which is widely used in mechanical system. However, bearing failure often affects mechanical operation and may lead to mechanical system paralysis when it is serious. According to the statistical data, the mechanical failure caused by bearing failure occupies a large proportion of the total mechanical failure. Therefore, the early fault monitoring of rolling bearing is carried out. It is of great significance to find out the location of the fault and predict the development direction of the fault. The key points in bearing fault diagnosis are to extract the bearing fault feature information from the signal to be processed, to judge the bearing fault type and fault location. Aiming at the problems of traditional morphology in fault signal processing of rolling bearing, such as difficult selection of structural elements and unsatisfactory noise reduction effect, the generalized difference filtering algorithm and adaptive Zhang Liang morphological algorithm are proposed. The main research contents of this paper are as follows:. 1. A bearing fault feature extraction algorithm based on generalized morphological difference filtering is proposed. According to the local feature information of the signal to be processed, the optimal structural element scale is selected to construct the generalized difference morphological filter. The fault characteristic information can be extracted better, and the method is proved to be effective by simulation signal and bearing simulation test signal. 2. On the basis of the generalized morphological difference filtering algorithm, in order to further improve the morphological noise reduction and feature extraction, An adaptive Zhang Liang morphology based bearing fault feature extraction algorithm is proposed in this paper. According to the local feature information of the signal to be processed, the elliptical structure element of Zhang Liang is constructed. Zhang Liang morphological filter is used to reduce noise and extract fault feature of bearing fault signal, which can replace the traditional linear structure element and disk structure element, and has a good application effect in bearing fault. 3. Comparing the advantages and disadvantages of the three kinds of morphological processing methods in the bearing fault feature extraction, the comprehensive analysis results show that the adaptive Zhang Liang morphology is the best in the feature extraction of the inner and outer ring faults of the bearing, the second is the generalized difference morphological extraction effect. The last one is the traditional morphology extraction effect, and the traditional morphology extraction effect is the best for the fault signal analysis of bearing rolling body, followed by the adaptive Zhang Liang morphology and the generalized difference morphology.
【学位授予单位】：武汉科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TH133.3;TH165.3

【参考文献】