基于盲源分离的旋转机械非平稳振动信号研究

发布时间：2018-04-17 02:22

本文选题：旋转机械 + 特征提取　；参考：《昆明理工大学》2011年硕士论文

【摘要】：旋转机械是工业生产部门中广泛应用的一类机械设备,是生产部门的核心设备,例如发电机、压缩机、汽轮机、鼓风机等机械设备都属于这一类。随着现代科学技术的迅猛发展和机械设备日趋向大型化、高速化、集成化、功能越来越多、结构越来越复杂、自动化程度越来越高,由此产生对旋转机械设备的管理以及维护要求也越来越高。此类设备能否正常运行,对生产部门和国民经济具有重要的意义。本文以旋转机械非平稳振动信号为研究对象,以盲源分离、包络分析和阶比分析为研究手段,研究对故障的提取和分离新途径,实现对齿轮故障和轴承故障特征提取以及同时存在两种故障的有效分离。论文分别对同时存在齿轮断齿故障和轴承外圈故障、内圈故障以及和滚动体故障的情况下进行了试验研究。提出了针对齿轮箱的多故障源分离方法,首先利用包络分析提取含有多种故障的特征信息的包络波形,然后结合独立分量分析,从混合信号中分离出信号中的独立源分量,最后在各独立分量中获得其对应的故障源,成功实现在复合故障下的故障特征提取与故障源的有效分离。研究中,在对独立分量分析技术进行了深入掌握的基础上,结合阶比包络谱分析技术,提出了“基于独立分量分析与包络阶比分析的齿轮箱多振源特征提取”方法。该方法解决了独立振源数目一般并不能预先确定,直接应用独立分量分析方法往往并不能实现对混合信号的有效分离问题,采用包络提取实现对原信号中振源数的降维,然后对包络波形进行阶比跟踪等角度采样,对等角度采样信号应用独立分量分析进行按源分离和包络阶比分析,提取出各振源的振动特征。研究表明该方法可实现对滚动轴承外圈故障和齿轮断齿故障的特征提取和分离,对旋转机械多源故障条件下特征提取技术的发展有较好的学术研究意义。
[Abstract]:Rotating machinery is a kind of machinery widely used in industrial production department. It is the core equipment of production department, such as generator, compressor, steam turbine, blower and so on.With the rapid development of modern science and technology and the increasing trend of machinery and equipment to large-scale, high-speed, integrated, more and more functions, more and more complex structure, the degree of automation is becoming higher and higher.As a result, the management and maintenance of rotating machinery are becoming more and more important.This kind of equipment can run normally, to production department and national economy have important meaning.In this paper, the non-stationary vibration signals of rotating machinery are taken as the research object, and the blind source separation, envelope analysis and order ratio analysis are taken as the research means to study the new ways of fault extraction and separation.The gear fault and bearing fault feature extraction are realized and the two faults are separated effectively.In this paper, the fault of gear tooth break and bearing outer ring, inner ring and rolling body are studied respectively.In this paper, a multi-fault source separation method for gearbox is proposed. Firstly, envelope analysis is used to extract the envelope waveform with the characteristic information of various faults, and then the independent source component is separated from the mixed signal by using the independent component analysis (ICA).Finally, the corresponding fault source is obtained in each independent component, and the fault feature extraction and the effective separation of fault source are successfully realized.In the research, on the basis of deeply mastering the independent component analysis technology and combining the order ratio envelope spectrum analysis technique, the method of "extracting the feature of multi-vibration source of gearbox based on independent component analysis and envelope order analysis" is put forward.This method solves the problem that the number of independent vibration sources can not be determined in advance, the direct application of independent component analysis method can not realize the effective separation of mixed signals, and the dimensionality reduction of the number of vibration sources in the original signal can be realized by using envelope extraction.Then the envelope waveform is sampled from different angles such as order tracking. The independent component analysis (ICA) is used to separate the envelope waveform and the envelope order ratio analysis is used to extract the vibration characteristics of each vibration source.The results show that this method can be used to extract and separate the features of the outer ring fault of rolling bearing and the fault of gear broken tooth, and it has a good academic significance for the development of feature extraction technology under the condition of multi-source fault of rotating machinery.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3

【参考文献】