摘要
为提高爆破震动与岩石破裂微震信号辨识精度,以某金属矿山现场微震监测数据为基础,首先采用小波包分解对矿山爆破震动和岩石破裂信号进行2层分解,对比两类信号在4个频段内的能量分布的特征,进而以两类事件的低频段(0～125 Hz)能量百分比、中低频段(125～250 Hz)量百分比、中高频段(250～375 Hz)能量百分比、高频段(375～500 Hz)能量百分比为特征向量,利用支持向量机(SVM)对爆破和岩石破裂微震信号进行了训练和分类,结果表明:两类事件在0～125 Hz的能量分布差异最大,且以0～125 Hz的能量百分比10%作为分界值时的准确率达到87.5%;SVM的分类正确率为94%,取得了理想的分类效果。
In order to improve identification accuracy of the blasting vibration and rock fracture micro-seismic signal, based on in-situ micro-seismic data,the micro-seismic signals were decomposed into 2 multi-scale,4 frequency bands to calculate the signals energy under different bands and their energy distribution is different. Then through using energy percent of low frequency(0~125Hz), medium-low frequency(125~250Hz), upper-middle(250~375Hz) frequency, high frequency(375~500Hz) as feature vector, SVM is adopted to train, classify the signals. Result shows that there are largest differences of energy distribution in low frequency(0~125Hz)between microseisms and blasts,and the best pattern recognition is obtained when energy percent of low frequency(0~125H)is 10% with an accuracy rate of 87.5%;The correct classified rate by SVM is 94%.
作者
杨晨
吴建星
Yang Chen;Wu Jianxing(College of Resource and Environment Engineering,Wuhan University of Science and Technology,Wuhan 430081,China)
出处
《科技通报》
2019年第1期19-23,共5页
Bulletin of Science and Technology
关键词
爆破震动与岩石破裂
小波包
SVM
频带能量
分类
blasting vibration and rock fracture
wavelet packet analysis
SVM
energy distribution
classifying
作者简介
杨晨(1993-),男,硕士研究生.,主要从事微震信号监测方面的研究。E-mail:wustyangchen@163.com。
