%0 Journal Article
%T 基于声发射波形参数的判别算法在模型试验中的应用研究<br>Application of Discrimination Algorithms Based on AE Waveform Parameters in Model Test
%A 孙强
%A 高淳
%A 袁海瑞
%A 宋银豪
%A 吴层层
%J Hans Journal of Civil Engineering
%P 1114-1124
%@ 2326-3466
%D 2019
%I Hans Publishing
%R 10.12677/HJCE.2019.86130
%X <div style=\"text-align:justify;\">
	本文以三次试验中单个孔隙类岩石单轴压缩全过程试验中获取的242万个声发射撞击事件为样本，基于持续时间、幅值和上升时间三个参数，应用判别算法获得八类声发射的特征波形，进而开展快速傅里叶变换(FFT)和matlab统计分析，重点研究岩石声发射信号的特征分类与应力的对应关系。研究表明：在整个试验阶段，八类类型波的频域主要集中在0~50 kHz和250~300 kHz二个频段，并随应力水平的变化其分布占比具有规律性；自始至终持续时间短、幅值较低、上升时间较长的波占据着大多数，接近峰值强度前，高幅值、上升时间较短的波形比例上升，持续时间长的波比例下降，进入破坏阶段后与之前相反。总体上波形参数中幅值与上升时间存在明显相关性对试件的破坏具有预警作用。该研究方法以整体大数据为研究基础，是对原有声发射参数评价方法的补充，后续进行不同类型岩石试验，将更加完善该方法的应用。<br />
Based on the duration, amplitude and rising time of 2.42 million acoustic emission impact events obtained in the whole process of uniaxial compression test of single porous rock in three tests, eight types of acoustic emission characteristic waveforms are obtained by discriminant algorithm, and then fast Fourier transform (FFT) and Matlab statistical analysis are carried out. This paper focuses on the relationship between the classification of rock acoustic emission signals and stress. The results show that the frequency domains of the eight types of waves are mainly concentrated in the two frequency bands of 0~50 kHz and 250~300 kHz, and their distribution proportion is regular with the change of stress level. Waves with short duration, low amplitude and long rise time occupy the majority. Before approaching the peak intensity, the proportion of waveforms with high amplitude and short rise time increases, while the proportion of waveforms with long duration decreases. After entering the destruction stage, the proportion of waveforms with high amplitude and short rise time decreases. In general, there is a significant correlation between the amplitude of the waveform parameters and the rising time, which has an early warning effect on the failure of the specimens. The research method is based on large data and is a supplement to the original acoustic emission parameter evaluation method. The subsequent rock tests of different types will improve the application of this method.
</div>
%K 模型试验，声发射，判别算法，波形参数，快速傅里叶变换(FFT)<br>Model Test
%K Acoustic Emission
%K Discriminant Algorithm
%K Waveform Parameters
%K Fast Fourier Transform (FFT)
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=31930