Study on Fractal Dimension about Vibration Signal Based on Wavelet Transform
Xiong Wei-hua · Journal of Zhejiang Sci-Tech University · 2011
Considering that the traditional methods are not obvious in judging the vibrations of the frequencies of the signal and difficult to extract fault feature,this paper,based on wavelet and fractal theory,by calculating box counting dimension,extracts vibration feature effectively.Wavelet fractal vibration signal high-frequency coefficient box counting dimension algorithm is proposed.Vibration signal is described in high-frequency.The experiment result shows that with the vibration of the frequency of the noise signal,there are more distinct vibration trend to high-frequency fractal dimension compared with the whole vibration signal fractal dimension and can better reflect the vibration of fault frequency.It has important pragmatic significance to vibration detection fault diagnosis.