Iterative generalized time-frequency reassignment for planetary gearbox fault diagnosis under nonstationary conditions
Iterative generalized time-frequency reassignment for planetary gearbox fault diagnosis under nonstationary conditions
复制标题
非平稳条件下行星齿轮箱故障诊断的迭代广义时频重分配
DOI：
10.1016/j.ymssp.2016.04.023
复制
发表时间：
2016-12-01
影响因子：
8.4
通讯作者：
Feng, Zhipeng
中科院分区：
文献类型：
--
作者：
Chen, Xiaowang;Feng, Zhipeng
行星齿轮箱以其结构紧凑、传动比大、承载能力强等优点，广泛应用于各种机械中。它们的故障诊断依赖于故障特征频率的有效识别。然而，除了复杂的机械运动学引起的振动复杂性之外，不稳定的外部条件导致时变的运行速度和/或负载，并且因此导致非平稳的振动信号。这往往导致故障特征时变复杂，给行星齿轮箱故障诊断增加了难度。时频分析是提取非平稳信号频率成分及其时变性的有效方法。然而，常用的时频分析方法存在时频分辨率低、内外干扰多等缺点，难以准确识别时变故障特征频率。时频重排虽然提高了时频的可读性，但它本质上受到瞬时频率的单分量和对称时频分布的约束。因此，它仍然容易受到错误的能量重新分配，甚至产生伪干扰，特别是对于高度非线性瞬时频率的多分量信号。针对时频重排的局限性，利用迭代广义解调的优点，提出了一种对非平稳多分量信号具有良好时频分辨率和抗干扰能力的改进方法。该方法首先通过迭代广义解调将信号分解为恒定频率的单分量;然后将时间-频率重新分配应用于每个广义解调的单分量，获得精细的时间-频率分布。最后，对各信号分量的时频分布进行恢复和叠加，得到原始信号的时频分布。利用数值模拟和实验室试验的行星齿轮箱振动信号对所提出的方法进行了验证。通过对时变齿轮故障征兆的提取，验证了迭代广义时频重分配方法在非平稳条件下行星齿轮箱故障诊断中的有效性。(C)2016爱思唯尔有限公司版权所有
Planetary gearboxes are widely used in many sorts of machinery, for its large transmission ratio and high load bearing capacity in a compact structure. Their fault diagnosis relies on effective identification of fault characteristic frequencies. However, in addition to the vibration complexity caused by intricate mechanical kinematics, volatile external conditions result in time-varying running speed and/or load, and therefore nonstationary vibration signals. This usually leads to time-varying complex fault characteristics, and adds difficulty to planetary gearbox fault diagnosis. Time-frequency analysis is an effective approach to extracting the frequency components and their time variation of nonstationary signals. Nevertheless, the commonly used time-frequency analysis methods suffer from poor time-frequency resolution as well as outer and inner interferences, which hinder accurate identification of time-varying fault characteristic frequencies. Although time frequency reassignment improves the time-frequency readability, it is essentially subject to the constraints of mono-component and symmetric time-frequency distribution about true instantaneous frequency. Hence, it is still susceptible to erroneous energy reallocation or even generates pseudo interferences, particularly for multi-component signals of highly nonlinear instantaneous frequency. In this paper, to overcome the limitations of time frequency reassignment, we propose an improvement with fine time-frequency resolution and free from interferences for highly nonstationary multi-component signals, by exploiting the merits of iterative generalized demodulation. The signal is firstly decomposed into mono-components of constant frequency by iterative generalized demodulation. Time-frequency reassignment is then applied to each generalized demodulated mono-component, obtaining a fine time-frequency distribution. Finally, the time-frequency distribution of each signal component is restored and superposed to get the time frequency distribution of original signal. The proposed method is validated using both numerical simulated and lab experimental planetary gearbox vibration signals. The time varying gear fault symptoms are successfully extracted, showing effectiveness of the proposed iterative generalized time-frequency reassignment method in planetary gearbox fault diagnosis under nonstationary conditions. (C) 2016 Elsevier Ltd. All rights reserved.