基于波形互相关的煤岩声发射事件自动识别模型

为了对煤岩破裂产生的大量声发射事件进行快速高效的自动识别,研究提出了一种基于长短时窗法(STA/LTA)的自适应滑动时窗更新方法(SSSW);根据煤岩破裂声发射波形相似性特征,在不断更新的自适应滑动时窗内构建了波形互相关函数,并采用加权最小二乘迭代算法对各通道波形进行时差校正;在时差校正的基础上,构建了声发射波形相似系数求解方程;综合上述研究,最终建立了基于波形互相关的煤岩声发射事件自动识别模型(WCCADM),WCCADM综合了各通道间有效声发射事件和噪声波形振幅、持续时间的相似性和差异性特征对煤岩破裂声发射事件进行自动识别。石灰岩胀裂声发射实验数据验证分析结果表明:SSSW可以根据煤岩破裂声发射事件发生时间间隔及波形持续时间不同的特征自适应地对滑动时窗长度进行调整更新;在自适应滑动时窗内,基于波形互相关函数、时差校正和相似系数方程得到各通道间波形的相似系数,通过设定合理的相似系数阈值能够对煤岩破裂诱发的大量声发射事件进行高效准确的自动识别。

Automatic detection model of acoustic emission events of coal and rock based on waveforms correlation

There occur many acoustic emission (AE) events during the crack process of coal and rock.In order to automatically detect these AE events fast and efficiently,an automatic detection model of acoustic emission events is constructed.Firstly,a self-adapted updating method of time windows based on STA/LTA method (SSSW) is put forward.Secondly,based on the similar characteristics of AE waveforms generated by coal and rock fracture,the correlation function of waveforms within the given time window is established.And then,through the weighted LS iterative algorithm,relative arrival time is given in order to align the AE waveforms.Thirdly,a function is developed to calculate the semblance coefficient and the AE events can be detected by comparing semblance coefficient value with the given threshold.Finally,the automatic detection model of acoustic emission events (WCCADM) is established based on the SSSW,the correlation function,and the semblance coefficient function.The WCCADM not only counts the signal noise ratio but also unifies the similarity and differences of amplitudes and wave durations between noise and AE events of coal and rock fracture.The operation process of and AE data generated by swelling fracture test of limestone indicated that:SSSW can determine the optimal length of time window according to the durations of waveforms and the time intervals between neighboring events.And the semblance coefficient given by the WCCADM can be used to detect the valid AE events among the massive quantity of AE data generated by coal and rock,which bases the research on AE source localization,focal mechanics and the fracture inversion both in time and space scales.