Study On Analysis Methods And Applications Of Blast Vibration Signals Based On The Wavelet Analysis

With the development of social economy and blasting technology application, people pay more and more attention to the harm of blasting vibration. But as for the collected blasting vibration signals, the data processing, there exist many deficiencies. Essentially speaking, blasting vibration is a non-stationary random process, wavelet transform's theory foundation and Fourier transform's are not uniform. After passing complex geotechnical medium, blasting seismic wave tend to have components of various frequency interference wave, and because of superposition of these interference wave, it brought many difficulties to the analysis of characteristics of blasting vibration signals. The traditional Fourier transform can only achieve to transform signal from the time domain to frequency domain, unable to realize the analysis of time domain and frequency domain of blasting vibration signals in the same time, wavelet analysis appear just offset the deficiency caused by Fourier transform to analyze blasting vibration signals.This dissertation introduces the basic theory of traditional Fourier analysis and wavelet analysis, processes Fourier spectrum analysis and wavelet time-frequency analysis to blasting vibration signal, including the following aspects:(1) expound the mathematical foundation and basic principle of wavelet analysis; the current blasting vibration signal analysis and processing methods are summarized; compare the traditional Fourier transform, short-time Fourier transform and gabor transformation; From the Angle of theory account for that wavelet analysis is in signal processing and analyzing advantages;(2) With the Blastmate SeriesⅢblasting vibration meter test, fast Fourier transform (FFT) analysis can be made directly. The physical quantities can be obtained from spectrum diagrams such as blasting vibration frequency of the signal in the circumstance of discretion, distribution and main vibration frequency. But, fast Fourier transform is also the conversion between time domain and frequency domain, the process of fast Fourier transformation from time domain to frequency domain, lost the concept of time, so that we can't get any time domain signal features from spectrum analysis, and the wavelet analysis exactly solved this problem. (3) For Containing noise or disturbance of measured blasting vibration signals, wavelet threshold denoising method for filter is proposed. Through comparing several threshold denoising method, the default threshold and soft threshold de-noising result is good that is found. With comparison of after Denoising blasting vibration signals and primitive blasting vibration signals, it was found that when high frequency information filter from the signal, vibration signal is becoming more and more "pure". blasting vibration signal has obvious contains no interference part through wavelet reconstruction, a real blast vibration signals was obtained, facilitate further to the signal wavelet analysis.(4) Introduces some of the most common time-frequency analysis method, short-time Fourier transform, gabor transform and wavelet transform. This paper introduced the basic theory of them, and uses them to analyze the measured blasting vibration signal. It was find through comparing that wavelet time-frequency analysis was more suitable than other two methods to do blasting vibration signals time-frequency analysis.(5) Application of wavelet analysis method, analyzes underground rock loose blasting vibration signal in coal mine. Wavelet analysis can realize the measured vibration signal of time, frequency, vibrating velocity three of the analysis of characteristics, to avoid the single time domain and frequency domain of Fourier analysis one-sidedness; Meanwhile, makes underground rock loose blasting vibration analysis in coal mine from blasting vibration the three aspects of three elements.(6) Application of the method that wavelet packet decomposition calculates different frequency band energy distribution, made the different frequency bands of energy distribution analysis to the data of measured vibration signal, obtained energy distribution characteristics of coal mine underground rock loose blasting vibration signal in different frequency bands.This paper through making the wavelet analysis to the measured vibration signal, and achieved good effect. Research shows that the wavelet analysis on the analysis of the blasting vibration signal has a big advantage. Believe that the wavelet analysis method in blasting vibration signal analysis will be more in-depth and more widely.