摘要
行星齿轮箱的瞬时频率(转速)精确提取是行星齿轮箱时变运行工况下故障诊断的关键。现有一些时频脊提取算法在提取瞬时频率时受到振动信号幅值变化以及时频变换结果中噪声成分的干扰,存在鲁棒性问题。为了解决这一问题,提出了一种双向搜索时频脊融合方法,并且将其进一步推广到多时频脊融合的概念中。然后,应用提出方法提取实际工程中时变工况行星齿轮箱高速轴的瞬时转速。再通过与现有两种方法对比以及阶次分析,验证了时频脊融合方法提取高速轴转速的可靠性与准确性。最后,在行星齿轮箱转速提取的基础上,利用阶次分析以及推导出的行星轮轴承故障频率的计算式,识别出了行星齿轮箱中的行星轮轴承内圈存在故障。
Planetary gearboxes often run under time-varying conditions, thus resulting in nonstationary signals. The accurate extraction of the reference shaft rotational speed is a key issue of planetary gearbox fault diagnosis. The conventional time-frequency ridge detection algorithms make a poor robustness for tracking the target time-frequency ridge due to the amplitude variation and noise buried in the raw vibration signal. In order to solve this problem, a bidirectional search time-frequency ridge fusion method is proposed, which could be further promoted into the fusion about the multiple time-frequency ridges. A vibration signal of planetary gearbox in practical engineering is used to validate the proposed method. As a result, the rotational speed of its high-speed shaft is effectively extracted by comparing with two conventional methods. Additionally, the order analysis is utilized to process the signal based on the extracted rotational speed. It is verified that the extracted rotational speed hasgood accuracy. Finally, with the expression of planet bearing faulty frequency, the inner race defect of the planet bearing is identified from the order spectrum.
出处
《振动工程学报》
EI
CSCD
北大核心
2017年第1期127-134,共8页
Journal of Vibration Engineering
基金
国家自然科学基金资助项目(51675262
51605319)
机械结构强度与振动国家重点实验室开放课题(SV2015-KF-01)
关键词
故障诊断
行星齿轮箱
时频脊
阶次分析
瞬时频率
fault diagnosis
planetary gearbox
time-frequency ridge
order analysis
instantaneous frequency
