Research On Noise Reduction And Pick-up Methods Of Underground Blasting Vibration Signal

Blasting operations in underground mining activities such as coal mining and subway construction can easily cause microseismic events.The real-time monitoring of microseismic events has been a hot topic for scholars at home and abroad.Accurately picking up the first arrival time of blasting vibration signal is the key link of the whole monitoring of microseismic events,and is also the premise and foundation of the location of blasting source.However,the underground blasting operation environment is complex,the blasting vibration signal collected by the sensor will be mixed with strong noise,and the useful characteristics of the signal are weak,so the traditional algorithm cannot accurately pick up the first arrival time of the signal.In order to meet the need of real-time monitoring and control for the location system of active blasting source in underground coal mine,it is difficult to pick up the first arrival of blasting vibration signal with low signal-to-noise ratio,in this paper,the denoising method of blasting vibration signal and the picking-up Algorithm of first break are deeply studied,which provides theoretical support for practical application.(1)A mathematical model is established based on the characteristics of fast attenuation and short duration of blasting vibration signal,and the wavelet threshold denoising algorithm is used to denoise the noise-containing blasting vibration signal.Subsequently,singular spectrum analysis is introduced into the choice of decomposition layers.According to the characteristic that the singular spectrum slope of wavelet decomposition coefficients tends to saturate with the increase of decomposition layers during the noise reduction process of blasting vibration signals,a self-single based on singular spectrum analysis is designed.And it realizes the adaptive selection of the optimal decomposition layers of blasting vibration signals.Moreover,in view of the shortcomings of the soft and hard threshold functions,a threshold function based on adjustment factors is designed,which can fulfil flexible adjustment between the soft and hard threshold functions by adjusting the size of parameter ?.At the same time,the threshold function can verify that the proposed threshold function has the best noise reduction effect through simulation.(2)For the purpose of obtaining accurate blasting vibration signal first arrival time,the advantages and disadvantages of STA/LTA(Short-Term to Long-Term Average)algorithm and AIC(Akaike information criterion)algorithm pick up first arrival time is analyzed.According to the characteristics that STA/LTA algorithm can roughly pick up the first arrival time of signal and AIC algorithm can pick up the result accurately but is easy to be limited by time window,S/L-AIC first arrival pick-up algorithm based on multi-feature function is proposed.However,the algorithm takes the minimum and maximum values of the first arrival and arrival time of the four feature functions of the STA/LTA algorithm as the reference,and expands the L/10 sampling points forward and backward according to the length L of the signal to form a new time window[min(ti)-L/10,max(ti)+L/10].Besides,the accurate first arrival time is picked under the time window,and it is verified that the proposed algorithm can accurately pick up the first arrival time of the blasting vibration signal under both high and low SNR conditions.(3)In order to verify the feasibility of S/L-AIC algorithm based on multi-feature function in the source location algorithm and the accuracy of first-arrival pickup,the advantages and disadvantages of five source location algorithms were analyzed.And the linear source location algorithm based on travel time theory was selected for simulation verification.By using MATLAB seismic simulation events,AIC algorithm and S/L-AIC algorithm based on nulti-feature function are used to pick up the first arrival of the simulated low SNR blasting vibration signal.According to the linear source positioning algorithm,the two kinds of algorithms are picked up.The source inversion location and precision analysis are carried out at the time of arrival.Consequently,the results verified the feasibility of the S/L-AIC algorithm based on multi-feature function in the source location and the high accuracy of the first arrival to the arrival.