摘要
分析了谐波小波的优点。在研究谐波小波的频段分解的基础上,提出了不分层分析的谐波窗方法。推导了谐波窗的频段分析表达式,并给出了实现过程。该方法显示了小波分析的窗口伸缩功能,但避免了在基于二进的小波变换过程中,由于隔点采样导致的信息丢失现象。利用该方法对3个数字信号进行分析,体现了该方法对近频信号和微弱瞬态信号的识别能力,也体现了该方法对淹没在强噪声中微弱周期信号的提取能力。利用该方法,对某汽车齿轮箱的裂纹故障产生的原因进行振动信号的频域识别,得到了满意结果。
This paper mines the advantage of harmonic wavelet analysis. On the bases of partial decomposition in frequency domain, a signal analysis method of harmonic windows that need not decomposition is presented. The equations of harmonic window analysis is proposed, and the implementation process is presented. It is shown that the width of the harmonic windows can be changed easily, but the method can avoid the phenomena of losing information in the dyadic wavelet transformation due to alternately sampling. By this method, three digital signals are analyzed. The identification ability is verified for the signals that the frequency is very close each other and for weak transient signals. It is also shown the pick-up ability of this method to weak periodic signal in the examples. Using the method, the reasons that extract creak fault in vibration signals of a automobile gearbox are analyzed. The result is satisfactory. This provides a new processing method for vibration signal of complex faults.
出处
《振动工程学报》
EI
CSCD
北大核心
2005年第2期252-256,共5页
Journal of Vibration Engineering
基金
航空科学基金资助项目(04152066)
山东省自然科学基金资助项目(Y2000A01)
关键词
故障诊断
微弱信号
裂纹故障
齿轮箱
谐波窗
小波变换
传动系统
Automobile parts and equipment
Failure (mechanical)
Frequency domain analysis
Signal processing
Vibrations (mechanical)
Wavelet transforms
