往复式压缩机多源冲击振动时频故障特征研究

发布时间：2018-10-07 19:45

【摘要】：往复压缩机结构复杂，激励源多，振动以多源冲击信号为主，表现出较强的非平稳、非高斯及非线性等复杂特性，采用传统振动信号分析方法无法有效提取往复压缩机故障信息，现已经成为往复压缩机故障诊断的瓶颈问题。振动信号的时频分析作为设备故障特征提取的先进手段，在旋转机械故障诊断过程中发挥着非常重要的作用。但是，随着故障诊断领域向着往复机械设备的延伸，传统的时频故障特征提取方法面临着挑战。针对往复压缩机复杂的振动信号，以短时傅氏变换、Wigner-Ville分布、小波变换及HHT变换等为基础的时频分析方法，都难以识别故障的有用信息，时频分布变得非常复杂，许多频率成分缺乏明确物理意义。本文针对往复式压缩机多源冲击振动特征，结合现代时频分析方法的优缺点，深入开展了往复式压缩机多源冲击振动时频故障特征研究。首先，采用小波变换、HHT等现代时频处理方法对调幅、调频及调频调幅等仿真信号进行了时频特征分析，并应用于往复式压缩机的故障诊断，虽然取得一些效果，但仍然存在分辨率低、边界效应干扰等问题，难以满足实际应用。然后，根据全局频率与瞬时频率两种极端性概念的局限性，提出了更具适用性的局部频率新概念，赋予多源冲击振动信号有明确物理意义的频率内涵，从这个概念出发，通过与自适应峰值分解方法结合，进一步与HHT时频分析方法进行仿真对比，结果表明，基于局部频率的时频分析方法更加有效揭示了多源冲击振动信号的特征，为往复压缩机故障诊断提供了一种新的手段。最后，对往复式压缩机的基本工作原理、主要部件的运动学及动力学进行分析，并指出了往复压缩机测试信号的强非平稳性，根据上述仿真分析成果，将基于EMD的局部频率时频分析方法及基于自适应峰值分解的广义局部频率时频分析方法应用于往复式压缩机故障诊断，取得较为显著的效果，，进一步验证了该方法的准确性和鲁棒性，为大型往复压缩机组故障诊断提供了更加丰富的故障特征信息。
[Abstract]:The reciprocating compressor has complex structure, many excitation sources, and the vibration is dominated by multi-source shock signals, showing strong complex characteristics such as non-stationary, non-Gao Si and nonlinear, etc. The traditional vibration signal analysis method can not effectively extract the fault information of reciprocating compressor, which has become the bottleneck problem in the fault diagnosis of reciprocating compressor. The time-frequency analysis of vibration signals, as an advanced method of equipment fault feature extraction, plays a very important role in the fault diagnosis of rotating machinery. However, with the extension of fault diagnosis to reciprocating machinery, the traditional time-frequency fault feature extraction method is facing a challenge. For the complex vibration signals of reciprocating compressors, the time-frequency analysis methods based on short time Fourier transform (STFT) Wigner-Ville distribution, wavelet transform and HHT transform are difficult to identify the useful information of faults, and the time-frequency distribution becomes very complex. Many frequency components lack clear physical meaning. Based on the characteristics of multi-source shock vibration of reciprocating compressor and the advantages and disadvantages of modern time-frequency analysis method, the time-frequency fault characteristics of multi-source shock vibration of reciprocating compressor are deeply studied in this paper. Firstly, the time-frequency characteristic analysis of the simulation signals such as amplitude modulation, frequency modulation and frequency modulation is carried out by using the modern time-frequency processing method such as wavelet transform and HHT, and applied to the fault diagnosis of reciprocating compressor, although some results have been obtained. However, there are still some problems such as low resolution, boundary effect interference and so on, so it is difficult to meet the practical application. Then, according to the limitation of the two extreme concepts of global frequency and instantaneous frequency, a more applicable new concept of local frequency is proposed, which endows the multi-source shock vibration signal with a clear physical meaning of frequency connotation. By combining with the adaptive peak decomposition method, the simulation results show that the time-frequency analysis method based on local frequency is more effective in revealing the characteristics of multi-source shock vibration signals than the HHT time-frequency analysis method. It provides a new method for fault diagnosis of reciprocating compressor. Finally, the basic working principle of reciprocating compressor, kinematics and dynamics of main components are analyzed, and the strong nonstationarity of test signal of reciprocating compressor is pointed out. The local frequency time-frequency analysis method based on EMD and the generalized local frequency time-frequency analysis method based on adaptive peak decomposition are applied to the fault diagnosis of reciprocating compressor. The accuracy and robustness of the method are further verified, and more abundant fault feature information is provided for fault diagnosis of large reciprocating compressor units.
【学位授予单位】：东北石油大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TH45;TH165.3

【相似文献】