地震研究

为解决用于高密度布设的低成本MEMS烈度计集成软、硬件资源有限,且难以嵌入较为复杂算法的这一问题,基于Matlab的仿真计算,通过讨论在不同特征函数、时窗长度和短窗位置下STA/LTA值的变化趋势、拾取效果和运算时间,以选取能提高算法灵敏性、改善地震事件拾取效果和提高算法计算效率的参数,并将改进的STA/LTA算法应用于实际地震数据处理。结果表明:不同的特征函数对事件拾取率、拾取时间偏差和算法运算时间影响不同; 长短时窗长度相差越大,STA/LTA值的变化越明显; 时窗越长,算法运算时间越长; 短窗置后可以增大STA/LTA值的变化幅度、减少算法计算量,改善算法拾取时间。改进的STA/LTA算法拾取效果更好,计算效率更高,占用内存资源更小,更适用于集成资源有限的MEMS烈度计。

Though a number of forecasting techniques are used for the detection of seismic events,each of them has its own pros and cons.Since the MEMS seismic intensity instrument which features low cost and follows the principle of high-density deployment has limitation in integrating software and hardware resources,it is difficult to embed some more complex algorithms.Based on the simulation calculation of Matlab,this paper aims to select the parameters that can improve algorithmic sensitivity,picking-effect of seismic events and calculation efficiency according to different factors of STA/LTA algorithm including variation tendency of the STA/LTA value,picking-effect and computing-time in various situations such as different characteristic functions,different window lengths and different short-window locations.And the improvement of STA/LTA algorithm is applied to practical seismic-data processing.The final analysis suggests that diverse characteristic functions have corresponding effects on facets like event picking-rate,picking-time deviation and computing-time of algorithms; the greater difference between the lengths of long-time windows and short-time ones,the more obviously the values of STA/LTA tend to change; the computing-time of algorithm depends on the length of the window(positive correlation); putting short windows after long ones can intensify variation range of STA/LTA values,reduce calculation amount and improve the picking time.In conclusion,the improved STA/LTA algorithm can improve the picking-effect,calculation efficiency and use less memory resources,so it is more suitable for MEMS intensity instruments with limited integrated resources.

引言
1 STA/LTA算法的基本原理
2 STA/LTA算法的参数分析
3 算法应用
4 结论

图1 不同频率下幅值由1变为5的模拟地震信号(a)和不同特征函数计算的STA/LTA值(b)<br/>Fig.1 The analog seismic signals with amplitude from 1 to 5(a)and the STA/LTA values(b)calculated by different characteristic funotions at different frequencies

图1 不同频率下幅值由1变为5的模拟地震信号(a)和不同特征函数计算的STA/LTA值(b)
Fig.1 The analog seismic signals with amplitude from 1 to 5(a)and the STA/LTA values(b)calculated by different characteristic funotions at different frequencies

表1 特征函数CF1(i)～CF4(i)的计算结果<br/>Tab.1 Calculation results of characteristic function CF1(i)-CF4(i)

表1 特征函数CF1(i)～CF4(i)的计算结果
Tab.1 Calculation results of characteristic function CF1(i)-CF4(i)

表2 不同特征函数的计算结果<br/>Tab.2 Calculation reswlts of different characteristic functions

表2 不同特征函数的计算结果
Tab.2 Calculation reswlts of different characteristic functions

图2 原始波形(a)以及不同长(b)、短(c)时窗长度计算得到的STA/LTA值<br/>Fig.2 Original waveforms(a)and STA/LTA values calculated with different long-time(b)and short-time(c)windows' lengths

图2 原始波形(a)以及不同长(b)、短(c)时窗长度计算得到的STA/LTA值
Fig.2 Original waveforms(a)and STA/LTA values calculated with different long-time(b)and short-time(c)windows' lengths

表3 相同时刻、不同时窗长度下STA/LTA值及算法运算时间<br/>Tab.3 STA/LTA values and operation time at the same time and in different time-window lengths

表3 相同时刻、不同时窗长度下STA/LTA值及算法运算时间
Tab.3 STA/LTA values and operation time at the same time and in different time-window lengths

图3 原始波形(a)和不同短窗位置计算的STA/LTA值(b)<br/>Fig.3 Onginal waveforms(a)and STA/LTA values calculated in different short-time window positions(b)

图3 原始波形(a)和不同短窗位置计算的STA/LTA值(b)
Fig.3 Onginal waveforms(a)and STA/LTA values calculated in different short-time window positions(b)

表4 改进STA/LTA算法与传统STA/LTA算法的拾取效果<br/>Tab.4 Picking effect of the traditional STA/LTA algorithm and the improved STA/LTA algorithm

表4 改进STA/LTA算法与传统STA/LTA算法的拾取效果
Tab.4 Picking effect of the traditional STA/LTA algorithm and the improved STA/LTA algorithm

表5 地震事件样本<br/>Tab.5 Samples of earthquake events

表5 地震事件样本
Tab.5 Samples of earthquake events

蔡莉,占伟伟,王秀,等.2014.基于微机电加速度计的地震烈度计[J].地震地磁观测与研究,35(1/2):240-246.
段建华,程建远,王云宏,等.2015.基于STA/LTA方法的微地震事件自动识别技术[J].煤田地质与勘探,43(1):76-80+85.
胡星星,高孟潭,滕云田,等.2015.地震动参数速报仪的研制[J].地球物理学报,58(3):1024-1034.
胡星星,滕云田,谢凡,等.2013.基于MEMS传感器的速度型地震计技术研究[J].地球物理学进展,28(1):515-522.
蒋策,吴建平,房立华.2018.地震检测与震相自动拾取研究[J].地震学报,40(1):45-57.
李昌珑.2013.基于MEMS加速度传感器的地震烈度监测技术研究与实现[D].北京:中国地震局地震研究所.
林建民,杨微,陈蒙,等.2012.偏振分析在地震信号检测中的应用[J].中国地震,28(2):133-143.
刘晗,张建中.2014.微震信号自动检测的STA/LTA算法及其改进分析[J].地球物理学进展,29(4):1708-1714.
刘劲松,王赟,姚振兴.2013.微地震信号到时自动拾取方法[J].地球物理学报,56(5):1660-1666.
刘希强,周蕙兰,郑治真,等.1998.基于小波变换的信号突变检测、滤波和瞬态谱研究[J].国际地震动态,(1):1-9.
马强,金星,李山有,等.2013.用于地震预警的P波震相到时自动拾取[J].地球物理学报,56(7):2313-2321.
牟培杰,王润秋,李彦鹏,等.2012.微地震事件的自动检测研究[C]//中国地球物理学会第二十八届年会论文集.北京:中国地球物理学会.
滕云田,王喜珍,王晓美,等.2006.用B-样条双正交小波拾取P波到时[J].地震学报,28(3):329-333.
王浩,丁炜.2013.MEMS加速度计与传统地震加速度计的比较研究[J].大地测量与地球动力学,33(S2):93-95.
王建军,吴荣辉,何加勇.2009.基于IPv6和无线网络的地震烈度计开发[J].现代电子技术,32(1):23-25,29.
王喜珍.2004.小波变换在地震数据压缩和震相到时拾取中的应用研究[D].北京:中国地震局地球物理研究所.
吴治涛,李仕雄.2010.STA/LTA算法拾取微地震事件P波到时对比研究[J].地球物理学进展,25(5):1577-1582.
余建华,李丹丹,韩国栋.2011.特征函数响应特性分析及STA/LTA方法的改进[J].国外电子测量技术,30(7):17-19,23,27.
张海涛,阎贵平.2003.MEMS加速度传感器的原理及分析[J].电子工艺技术,(6):260-262,265.
赵岑.2013.P波预警中的震级预测和PGA、PGV估算研究[D].成都:西南交通大学.
赵大鹏,刘希强,李红,等.2012.峰度和AIC方法在区域地震事件和直达P波初动自动识别方面的应用[J].地震研究,35(2):220-225.
Akaike H.1973.Information theory as an extension of the maximum likelihood principle[C]// Proceedings in second international symposium on information theory.Budapest Akademiai Kiado,267-281.
Allen R E.1978.Automatic earthquake recognition and timing form single traces[J].Bulletin of the Seismological Research of America,68(5):1521-1532.
Bai C Y.2000.Automatic phase-detection and identification by full Use of a single three-component broadband seismogram[J].Bulletin of the Seismological Society of America,90(1):187-198.
Chen Y N,Li J,Wang Z J,et al.2020.Quick phase identification for dense seismic array with aid from long term phase records of co-located sparse permanent stations[J].Earthquake Research in China,34(3):328-342.
Jubran A,David E.2012.Adaptive microseismic event detection and automatic time picking[C].Geo Convention 2012:Vision,1-5.
Kwon J,Heo T,Kim J K,et al.2018.A new P-wave detector via moving empirical cumulative distribution function[C].Bulletin of the Seismological Society of America,108(4):2080-2089.
Massan,Ferrettig,Spallarossad,et al.2006.Improving automatic location procedure by waveform similarity analysis:An application in the South Western Alps(Italy)[J].Play Earth Planet Interiors,154:18-29.
Peng C,Chen Y,Chen Q,et al.2017.A new type of tri-axial accelerometers with high dynamic range MEMS for earthquake early warning[J].Computers & Geosciences,100(C):179-187.
Peng C,Jiang P,Chen Q,et al.2019.Performance evaluation of a dense MEMS-based seismic sensor array deployed in the sichuan-Yunnan border region for earthquake early warning[J].Micromachines,10(11):735.
Ruud B O,Huserbye E S.1992.A new three-component detector and automatic single-station bulletin production[J].Bulletin of the Seismological Research of America,82(1):221-237.
Sleeman R,Eck T V.1999.Robust automatic P-phase picking:an on-line implementation in the analysis of broadband seismogram recordings[J].Physics of the Earth and Planetary Interiors,113(1-4):265-275.
Stevenson R.1976.Microearthquakes at Flathead Lake,Montana:A study using automatic earthquake processing[J].Bulletin of the Seismological Research of America,66(5):61-79.
Takanami T,Kitagawa G.1993.Multivariate time-series model to estimate the arrival times of S-waves[J].Pergamon,19(2):295-301.
Xu Z,Wang T,Xu S H,et al.2019.Active source seismic identification and automatic picking of the P-wave first arrival using a convolutional neural network[J].Earthquake Research in China,33(2):288-304.

备注

引言

1 STA/LTA算法的基本原理

2 STA/LTA算法的参数分析

3 算法应用

4 结论

期刊信息