基于小波变换的矿井提升机钢丝绳故障识别研究

发布时间：2018-03-25 11:42

本文选题：钢丝绳断丝　切入点：漏磁检测　出处：《中国矿业大学》2016年硕士论文

【摘要】：钢丝绳由于自身重量轻、弹性好、抗拉强度和疲劳强度高、工作平稳可靠、承受过载能力强以及在高速运行条件下卷扬噪声小等优点,在矿产、冶金、交通、建筑、旅游等领域获得了广泛的应用,是煤矿生产中极为重要的一环。然而其作为承载部件,在使用过程中不可避免的会产生类似锈蚀、局部断丝、磨损的缺陷,这些小缺陷逐渐积累即会造成钢丝绳的整根断裂,严重影响煤矿安全生产。因此钢丝绳的故障识别研究对保护煤矿的人身和财产安全具有重要的意义。本文分析了钢丝绳的基本结构和分类以及钢丝绳故障检测的标准,并根据钢丝绳结构选取了故障漏磁检测方法。设计了故障漏磁检测中励磁装置的结构并选择了合适的励磁强度。通过建立钢丝绳故障的等效点偶极子和等效带偶极子两种模型对钢丝绳故障进行数值模拟,分析各个参数对钢丝绳故障检测所造成的影响,并对两种模型进行对比和分析。同时利用ANSYS软件对故障钢丝绳进行磁场有限元分析,得到钢丝绳故障附近的漏磁场分布规律。分析钢丝绳故障诊断中可能会引入的噪声信息,并对这些不同类型的噪声进行特征的归纳。在传统带阻滤波器的基础上利用频域特性设计改进的自适应多重带阻滤波进行股波噪声的剔除。鉴于经典小波的频带混叠以及平移不变性差的特点,引入双树复小波进行钢丝绳故障数据的预处理和分析,并对比经典小波和双树复小波在频带混叠以及平移不变性上性能的差异。最后利用奇异性分析以及多尺度分析等算法对钢丝绳漏磁场仿真信号进行故障识别。为了验证算法的可行性,搭建了钢丝绳故障检测实验台,利用Lab VIEW编写信号采集软件完成内部断丝及外部断丝的数据采集。利用改进的自适应多重带阻滤波、双树复小波变换、奇异性分析以及120~128尺度的小波分析进行实验数据的预处理和故障识别。分析不同根数的内部断丝以及外部断丝故障趋势的异同点,并在故障诊断过程中成功检测到轻度磨损这一未预期的缺陷,证明了漏磁检测方法以及本文算法的准确性和可行性。
[Abstract]:Because of its light weight, good elasticity, high tensile strength and fatigue strength, stable and reliable work, strong overloading capacity and low lifting noise under high-speed operation conditions, steel wire ropes are used in minerals, metallurgy, traffic, construction, etc. Tourism and other fields have been widely used, which is an extremely important link in coal mine production. However, as a bearing part, it will inevitably produce defects such as rust, partial broken wire and wear in the process of use. The accumulation of these small defects will cause the whole wire rope to break. Therefore, it is of great significance to study the fault identification of wire rope to protect the safety of person and property in coal mine. This paper analyzes the basic structure and classification of wire rope and the standard of wire rope fault detection. According to the structure of wire rope, the fault magnetic leakage detection method is selected. The structure of excitation device in fault magnetic leakage detection is designed and the appropriate excitation intensity is selected. The equivalent point dipole and equivalent band pair of wire rope fault are established. Two kinds of pole models are used to simulate the fault of wire rope. The influence of each parameter on the fault detection of wire rope is analyzed, and the two models are compared and analyzed. At the same time, the finite element analysis of the magnetic field of the faulty wire rope is carried out by using ANSYS software. The distribution law of magnetic field leakage near the fault of wire rope is obtained. The noise information that may be introduced in fault diagnosis of wire rope is analyzed. Based on the traditional band-stop filter, an improved adaptive multi-band-stop filter in frequency domain is used to eliminate the noise of the femoral wave. In view of the frequency band of the classical wavelet, the characteristics of these different types of noise are summarized. The characteristics of aliasing and translation invariance, The double tree complex wavelet is introduced to preprocess and analyze the wire rope fault data. The difference between classical wavelet and double tree complex wavelet in frequency band aliasing and translation invariance is compared. Finally, the fault identification of wire rope magnetic field leakage simulation signal is carried out by using singularity analysis and multi-scale analysis. The feasibility of the algorithm is verified. In this paper, the fault detection test bench of wire rope is built, and the data acquisition of internal and external broken wire is accomplished by using Lab VIEW. The improved adaptive multi-band stop filter and double tree complex wavelet transform are used. The singularity analysis and the wavelet analysis of 120-128 scale are used to preprocess the experimental data and identify the faults. The similarities and differences of internal and external breakage trends of different root numbers are analyzed. The unexpected defect of mild wear was successfully detected in the process of fault diagnosis, which proved the accuracy and feasibility of the method of magnetic flux leakage detection and the algorithm in this paper.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TD532

【相似文献】