发动机振动信号的时-频域二次去噪方法
Time-frequency Domain Secondary Denoising Method for Engine Vibration Signals
陈曦明 1王忠 1葛铠 1黄印玉 1高永利 1肖鹏 1吴建洋1
作者信息
- 1. 北京航天发射技术研究所,北京 100076
- 折叠
摘要
本文提出一种适用于发动机机体振动信号的时频域二次去噪方法,从时域和频域两个方向对机体振动信号进行两次去噪.其中时域去噪过程首先是利用包络导数算子和短时能量概念来对时域上的噪声信号和工作特征信号进行识别和分割,再通过非局部均值(Non-Local Mean,NLM)方法实现去噪.而频域去噪过程则是先基于时变滤波-经验模态分解(Time Varying Filtering-Empirical Mode Decomposition,TVF-EMD)对振动信号进行时频特征提取,再提出一种融合差分熵的NLM方法对提取的本征模函数进行自适应去噪.应用所提出的去噪方法对实测的发动机机体振动信号进行分析,结果表明:该方法在时频域的去噪效果十分显著,在发动机状态监测和故障特征提取方面具有较大的应用潜力.
Abstract
This paper proposed a time-frequency domain quadratic denoising method for engine block vibration signals.The vibration signals will be denoised twice from time domain and frequency domain.In the time domain denoising process,the envelope derivative operator and the concept of short-time energy were first used to identi-fy and segment the noise signal and the working feature signal in the time domain,and then the Non-Local Mean(NLM)method was used to achieve denoising.In the frequency domain denoising process,the vibration signal was extracted based on Time Varying Filtering-Empirical Mode Decomposition(TVF-EMD).Then a new NLM method with differential entropy was proposed for adaptive denoising of the extracted intrinsic mode functions.The proposed denoising method was used to analyze the measured vibration signals of diesel engine body.The re-sults showed that the method has remarkable denoising effect in time-frequency domain,and has great application potential in diesel engine condition monitoring and fault feature extraction.
关键词
发动机/振动信号去噪/时频特征/NLM/TVF-EMDKey words
Engine/Vibration signal denoising/Time-frequency characteristics/NLM/TVF-EMD引用本文复制引用
基金项目
国家自然科学青年基金项目(52202457)
出版年
2024
NETL
NSTL
万方数据