排序方式： 共有75条查询结果，搜索用时 46 毫秒
51.
Automatic picking of P and S phases using a neural tree 总被引：2，自引：1，他引：2
The large amount of digital data recorded by permanent and temporary seismic networks makes automatic analysis of seismograms and automatic wave onset time picking schemes of great importance for timely and accurate event locations. We propose a fast and efficient P- and S-wave onset time, automatic detection method based on neural networks. The neural networks adopted here are particular neural trees, called IUANT2, characterized by a high generalization capability. Comparison between neural network automatic onset picking and standard, manual methods, shows that the technique presented here is generally robust and that it is capable to correctly identify phase-types while providing estimates of their accuracies. In addition, the automatic post processing method applied here can remove the ambiguity deriving from the incorrect association of events occurring closely in time. We have tested the methodology against standard STA/LTA phase picks and found that this neural approach performs better especially for low signal-to-noise ratios. We adopt the recall, precision and accuracy estimators to appraise objectively the results and compare them with those obtained with other methodologies.Tests of the proposed method are presented for 342 earthquakes recorded by 23 different stations (about 5000 traces). Our results show that the distribution of the differences between manual and automatic picking has a standard deviation of 0.064 s and 0.11 s for the P and the S waves, respectively. Our results show also that the number of false alarms deriving from incorrect detection is small and, thus, that the method is inherently robust.This paper has not been submitted elsewhere in identical or similar form, nor will it be during the first three months after its submission to Journal of Seismology. 相似文献
52.
The seismic wave consists of many seismic phases, which contain rich geophysical information from the hypocenter, medium of seismic wave passing through and so on. It is very important to detect and pick these seis-mic phases for understanding the mechanism of earthquake, the Earth structure and property of seismic waves. In order to reduce or avoid the loss resulted from the earthquake, one of the important goals of seismic event detect-ing is to obtain its related information before and afte… 相似文献
53.
Jaromír Janský Vladimír Plicka Helene Lyon-Caen Oldřich Novotný 《Journal of Seismology》2007,11(2):199-204
We determine the velocities in an upper crustal model, composed of three homogeneous layers, for one subregion of the western
part of the Gulf of Corinth, NE of the town of Aigion, Greece. We have used local events that occurred there in the year 2001
and were recorded by the Corinth Rift Laboratory Network. Weighted P and S arrival time residuals are minimized using the Neighbourhood Algorithm of Sambridge (1999), combined with the grid search
for source locations. The resolution of the inversion is tested by delete-one jackknifing. The model obtained is compared
with some other models derived or applied to the subregion. A fast velocity increase between depths of 5 and 7 km is confirmed
as the major structural element. 相似文献
54.
以JOPENS系统实时流接收为基础,应用Redis共享内存技术和近年来发展较快的深度学习震相自动识别技术,设计一套可7×24小时不间断稳定接收并实时识别连续地震流数据中P、S震相的系统,为地震台网实时数据处理提供一套辅助工具,并在福建省地震局测震台网128个台站的实时数据流上进行测试。该工具由Redis实时数据流共享模块与深度学习震相到时自动拾取、MSDP震相格式转换3个模块组成,可以实时接收并自动识别台网地震连续波形,生成P、S震相报告,并可导入MSDP人机交互工具进一步处理,在一定程度上可以减轻人工处理工作量。 相似文献
55.
基于分形理论,提出了一种地震波初至的拾取方法,根据地震初至波到达前后时间序列的分形特征,采用适当的时间窗口,可以定量地确定初至波的到达时间。数值计算表明,当窗口大小为0.6 s时,计算速度合适,可以保证地震波波至时间的拾取精度,并且在广西地区的宽频带地震流动台网记录的地震数据处理中进行了应用。 相似文献
56.
Emil Blias 《Geophysical Prospecting》2012,60(6):1024-1029
A new, adaptive multi‐criteria method for accurate estimation of three‐component three‐dimensional vertical seismic profiling of first breaks is proposed. Initially, we manually pick first breaks for the first gather of the three‐dimensional borehole set and adjust several coefficients to approximate the first breaks wave‐shape parameters. We then predict the first breaks for the next source point using the previous one, assuming the same average velocity. We follow this by calculating an objective function for a moving trace window to minimize it with respect to time shift and slope. This function combines four main properties that characterize first breaks on three‐component borehole data: linear polarization, signal/noise ratio, similarity in wave shapes for close shots and their stability in the time interval after the first break. We then adjust the coefficients by combining current and previous values. This approach uses adaptive parameters to follow smooth wave‐shape changes. Finally, we average the first breaks after they are determined in the overlapping windows. The method utilizes three components to calculate the objective function for the direct compressional wave projection. An adaptive multi‐criteria optimization approach with multi three‐component traces makes this method very robust, even for data contaminated with high noise. An example using actual data demonstrates the stability of this method. 相似文献
57.
58.
59.
震相到时的自动精确检测对实现海量波形数据自动处理有重要意义. 针对流动地震台阵观测,本文综合利用单台Akaike信息准则（AIC）和多台最小二乘互相关方法,发展了震相自动精确检测技术. 检测结果表明,在长短时平均比值方法（STA/LTA）检测地震事件的基础上,利用单台AIC方法,近震初至震相检测精度小于0.3 s;利用多台最小二乘互相关方法,能够可靠地检测高信噪比地震的初至震相到时,当信噪比较低时,能够有效地避免初至震相的错误判别. 相似文献
60.
地震震相拾取是地震数据自动处理的首要环节,包括了信号检测、到时估计和震相识别等过程,震相拾取的准确性直接影响到后续事件关联处理的性能,影响观测报告的质量.为了提高震相拾取的准确性,进而提高观测报告质量,本文采用深度卷积神经网络方法来解决震相拾取问题,构建了多任务卷积神经网络模型,设计了分类和回归的联合损失函数,定义了基于加权的分类损失函数,以三分量地震台站的波形数据作为输入,同时实现对震相的检测识别和到时的精确估计.利用美国南加州地震台网的200万条震相和噪声数据对模型进行训练、验证和测试,对于测试集中直达波P、S震相识别的查全率达到98%以上,到时估计的标准偏差分别为0.067s,0.082s.利用迁移学习和数据增强,将模型用于对我国东北地区台网的6个台站13000条数据的训练、验证和测试中,对该数据集P、S震相查全率分别达到91.21%、85.65%.基于迁移训练后的模型,设计了用于连续数据的震相拾取方法,利用连续的地震数据对该算法进行了实际应用测试,并与国家数据中心和中国地震局的观测报告进行比对,该方法的震相检测识别率平均可达84.5%,验证了该方法在实际应用中的有效性.本文所提出的方法展示了深度神经网络在地震震相拾取中的优异性能,为地震震相和事件的检测识别提供了新的思路. 相似文献