Fast and accurate P-wave arrival picking significantly affects the performance of earthquake early warning(EEW)systems.Automated P-wave picking algorithms used in EEW have encountered problems of falsely picking up no...Fast and accurate P-wave arrival picking significantly affects the performance of earthquake early warning(EEW)systems.Automated P-wave picking algorithms used in EEW have encountered problems of falsely picking up noise,missing P-waves and inaccurate P-wave arrival estimation.To address these issues,an automatic algorithm based on the convolution neural network(DPick)was developed,and trained with a moderate number of data sets of 17,717 accelerograms.Compared to the widely used approach of the short-term average/long-term average of signal characteristic function(STA/LTA),DPick is 1.6 times less likely to detect noise as a P-wave,and 76 times less likely to miss P-waves.In terms of estimating P-wave arrival time,when the detection task is completed within 1 s,DPick′s detection occurrence is 7.4 times that of STA/LTA in the 0.05 s error band,and 1.6 times when the error band is 0.10 s.This verified that the proposed method has the potential for wide applications in EEW.展开更多
In seismic data processing,picking of the P-wave first arrivals takes up plenty of time and labor,and its accuracy plays a key role in imaging seismic structures.Based on the convolution neural network(CNN),we propose...In seismic data processing,picking of the P-wave first arrivals takes up plenty of time and labor,and its accuracy plays a key role in imaging seismic structures.Based on the convolution neural network(CNN),we propose a new method to pick up the P-wave first arrivals automatically.Emitted from MINI28 vibroseis in the Jingdezhen seismic experiment,the vertical component of seismic waveforms recorded by EPS 32-bit portable seismometers are used for manually picking up the first arrivals(a total of 7242).Based on these arrivals,we establish the training and testing sets,including 25,290 event samples and 710,616 noise samples(length of each sample:2 s).After 3,000 steps of training,we obtain a convergent CNN model,which can automatically classify seismic events and noise samples with high accuracy(>99%).With the trained CNN model,we scan continuous seismic records and take the maximum output(probability of a seismic event)as the P-wave first arrival time.Compared with STA/LTA(short time average/long time average),our method shows higher precision and stronger anti-noise ability,especially with the low SNR seismic data.This CNN method is of great significance for promoting the intellectualization of seismic data processing,improving the resolution of seismic imaging,and promoting the joint inversion of active and passive sources.展开更多
Accurately picking P-and S-wave arrivals of microseismic(MS)signals in real-time directly influences the early warning of rock mass failure.A common contradiction between accuracy and computation exists in the current...Accurately picking P-and S-wave arrivals of microseismic(MS)signals in real-time directly influences the early warning of rock mass failure.A common contradiction between accuracy and computation exists in the current arrival picking methods.Thus,a real-time arrival picking method of MS signals is constructed based on a convolutional-recurrent neural network(CRNN).This method fully utilizes the advantages of convolutional layers and gated recurrent units(GRU)in extracting short-and long-term features,in order to create a precise and lightweight arrival picking structure.Then,the synthetic signals with field noises are used to evaluate the hyperparameters of the CRNN model and obtain an optimal CRNN model.The actual operation on various devices indicates that compared with the U-Net method,the CRNN method achieves faster arrival picking with less performance consumption.An application of large underground caverns in the Yebatan hydropower station(YBT)project shows that compared with the short-term average/long-term average(STA/LTA),Akaike information criterion(AIC)and U-Net methods,the CRNN method has the highest accuracy within four sampling points,which is 87.44%for P-wave and 91.29%for S-wave,respectively.The sum of mean absolute errors(MAESUM)of the CRNN method is 4.22 sampling points,which is lower than that of the other methods.Among the four methods,the MS sources location calculated based on the CRNN method shows the best consistency with the actual failure,which occurs at the junction of the shaft and the second gallery.Thus,the proposed method can pick up P-and S-arrival accurately and rapidly,providing a reference for rock failure analysis and evaluation in engineering applications.展开更多
Microseismic phase detection and arrival picking are critical steps in the processing of hydraulic fracturing microseismic monitoring data.As the signal-to-noise ratios of P-and S-waves diff er because of the infl uen...Microseismic phase detection and arrival picking are critical steps in the processing of hydraulic fracturing microseismic monitoring data.As the signal-to-noise ratios of P-and S-waves diff er because of the infl uence of focal mechanisms,geometry,and background noise,diffi culties are introduced in the eff ective discrimination of seismic phases and the accurate acquisition of arrivals in conventional processing methods.In this paper,we propose a method for identifying microseismic phase arrival time by comprehensively analyzing the variation of moveout curves and combining the intra-event waveform similarities with the energy ratio of multitrace signals.First,a curve-fi tting formula is constructed with perforation arrivals,and event detection is achieved by adopting an energy-weighted similarity coeffi cient that seeks plausible fi tting curves with a sliding time window in continuous microseismic recordings.Then,the P-and S-waves are separated by the fitting parameters.The known arrival time trend of the microseismic phase is employed to calculate residual time corrections.Finally,the accurate arrival results of the microseismic phases can be obtained by picking the arrivals of stacked traces.The reliability and eff ectiveness of the proposed method for microseismic phase detection and arrival picking were determined through tests using field data.Arrival results indicate that the proposed method can improve accuracy compared with the traditional energy ratio method.展开更多
Accurately detecting the arrival time of a channel wave in a coal seam is very important for in-seam seismic data processing. The arrival time greatly affects the accuracy of the channel wave inversion and the compute...Accurately detecting the arrival time of a channel wave in a coal seam is very important for in-seam seismic data processing. The arrival time greatly affects the accuracy of the channel wave inversion and the computed tomography （CT） result. However, because the signal-to-noise ratio of in-seam seismic data is reduced by the long wavelength and strong frequency dispersion, accurately timing the arrival of channel waves is extremely difficult. For this purpose, we propose a method that automatically picks up the arrival time of channel waves based on multi-channel constraints. We first estimate the Jaccard similarity coefficient of two ray paths, then apply it as a weight coefficient for stacking the multi- channel dispersion spectra. The reasonableness and effectiveness of the proposed method is verified in an actual data application. Most importantly, the method increases the degree of automation and the pickup precision of the channel-wave arrival time.展开更多
基于微机电系统(Micro-Electro-Mechanical System,MEMS)技术研制的MEMS强震仪具有易集成、维护成本低和低功耗等优点,在地震监测领域应用广泛.然而,MEMS强震仪集成的软、硬件资源有限,并且受仪器自身噪声等因素干扰较大,地震信号测量...基于微机电系统(Micro-Electro-Mechanical System,MEMS)技术研制的MEMS强震仪具有易集成、维护成本低和低功耗等优点,在地震监测领域应用广泛.然而,MEMS强震仪集成的软、硬件资源有限,并且受仪器自身噪声等因素干扰较大,地震信号测量结果质量较低,对嵌入算法要求更高.针对这一问题,本文提出一种更适用于MEMS强震仪的改进长短时窗均值比(Short Term Average/Long Term Average,STA/LTA)算法.首先,通过构建抗干扰(Anti-interference,AR)特征函数抑制基线漂移和低频噪声的干扰,提高STA/LTA算法拾取地震事件的抗干扰能力;其次,提出采用“延时长窗”的方式,提高STA/LTA算法的计算效率和拾取精度,减少STA/LTA算法对MEMS集成资源的占用;最后,结合时窗位置进一步探究不同时窗大小对STA/LTA算法拾取效率的影响.实际地震资料处理结果表明,本文提出的改进STA/LTA算法计算效率更高,实时性和抗干扰能力更强,更适用于集成资源有限的MEMS强震仪.展开更多
Passive microseismic monitoring(PMM)serves as a fundamental technology for assessing hydraulic fracturing(HF)effectiveness,with a key focus on accurate and efficient phase detection/arrival picking and source location...Passive microseismic monitoring(PMM)serves as a fundamental technology for assessing hydraulic fracturing(HF)effectiveness,with a key focus on accurate and efficient phase detection/arrival picking and source location.In PMM data processing,the data-driven paradigm(deep learning based)outperforms the model-driven paradigm in characteristic extraction but lacks quality control and uncertainty quantification.Monte Carlo Dropout,a Bayesian uncertainty quantification technique,performs stochastic neuron deactivation through multiple forward propagation samplings.Therefore,this study proposes a deep learning neural network incorporating uncertainty quantification with manual quality control integration,establishing an optimized workflow spanning automated phase detection to robust source location.The methodology implementation comprises two principal components:(1)The MDNet employing Monte Carlo Dropout strategy enabling simultaneous phase detection/arrival picking and unce rtainty estimation;(2)an integrated hybrid-driven workflow with a traveltime-based inve rsion method for source location.Validation with field data demonstrates that MD-Net achieves superior performance under low signal-to-noise ratio conditions,maintaining detection accuracy exceeding 99%for both P-and S-waves.The phase arrival picking precision shows significant improvement,with a 40%reduction in standard deviation compared to the baseline model(P-S time difference decreasing from12.0 ms to 7.1 ms),while providing quantifiable uncertainty metrics for manual calibration.Source location results further reveal that our hybrid-driven workflow produces more physically plausible event distributions,with 100%of microseismic eve nts clustering along the primary fracture expanding direction.This performance surpasses traditional cross-correlation methods and single/multi-trace data-driven me thods in spatial rationality.This study establishes an inte rpretable,high-pre cision automated framework for HF-PMM applications,demonstrating potential for extension to diverse geological settings and monitoring configurations.展开更多
基金National Natural Science Foundation of China under Grant Nos.51968016 and 5197083806the Guangxi Innovation Driven Development Project(Science and Technology Major Project,Grant No.Guike AA18118008).
文摘Fast and accurate P-wave arrival picking significantly affects the performance of earthquake early warning(EEW)systems.Automated P-wave picking algorithms used in EEW have encountered problems of falsely picking up noise,missing P-waves and inaccurate P-wave arrival estimation.To address these issues,an automatic algorithm based on the convolution neural network(DPick)was developed,and trained with a moderate number of data sets of 17,717 accelerograms.Compared to the widely used approach of the short-term average/long-term average of signal characteristic function(STA/LTA),DPick is 1.6 times less likely to detect noise as a P-wave,and 76 times less likely to miss P-waves.In terms of estimating P-wave arrival time,when the detection task is completed within 1 s,DPick′s detection occurrence is 7.4 times that of STA/LTA in the 0.05 s error band,and 1.6 times when the error band is 0.10 s.This verified that the proposed method has the potential for wide applications in EEW.
基金sponsored by the National Key Research and Development Project(2018YFC1503202-01)the Emergency Management Project of the National Natural Science Foundation of China(41842042)
文摘In seismic data processing,picking of the P-wave first arrivals takes up plenty of time and labor,and its accuracy plays a key role in imaging seismic structures.Based on the convolution neural network(CNN),we propose a new method to pick up the P-wave first arrivals automatically.Emitted from MINI28 vibroseis in the Jingdezhen seismic experiment,the vertical component of seismic waveforms recorded by EPS 32-bit portable seismometers are used for manually picking up the first arrivals(a total of 7242).Based on these arrivals,we establish the training and testing sets,including 25,290 event samples and 710,616 noise samples(length of each sample:2 s).After 3,000 steps of training,we obtain a convergent CNN model,which can automatically classify seismic events and noise samples with high accuracy(>99%).With the trained CNN model,we scan continuous seismic records and take the maximum output(probability of a seismic event)as the P-wave first arrival time.Compared with STA/LTA(short time average/long time average),our method shows higher precision and stronger anti-noise ability,especially with the low SNR seismic data.This CNN method is of great significance for promoting the intellectualization of seismic data processing,improving the resolution of seismic imaging,and promoting the joint inversion of active and passive sources.
基金We acknowledge the funding support from National Natural Science Foundation of China(Grant No.42077263).
文摘Accurately picking P-and S-wave arrivals of microseismic(MS)signals in real-time directly influences the early warning of rock mass failure.A common contradiction between accuracy and computation exists in the current arrival picking methods.Thus,a real-time arrival picking method of MS signals is constructed based on a convolutional-recurrent neural network(CRNN).This method fully utilizes the advantages of convolutional layers and gated recurrent units(GRU)in extracting short-and long-term features,in order to create a precise and lightweight arrival picking structure.Then,the synthetic signals with field noises are used to evaluate the hyperparameters of the CRNN model and obtain an optimal CRNN model.The actual operation on various devices indicates that compared with the U-Net method,the CRNN method achieves faster arrival picking with less performance consumption.An application of large underground caverns in the Yebatan hydropower station(YBT)project shows that compared with the short-term average/long-term average(STA/LTA),Akaike information criterion(AIC)and U-Net methods,the CRNN method has the highest accuracy within four sampling points,which is 87.44%for P-wave and 91.29%for S-wave,respectively.The sum of mean absolute errors(MAESUM)of the CRNN method is 4.22 sampling points,which is lower than that of the other methods.Among the four methods,the MS sources location calculated based on the CRNN method shows the best consistency with the actual failure,which occurs at the junction of the shaft and the second gallery.Thus,the proposed method can pick up P-and S-arrival accurately and rapidly,providing a reference for rock failure analysis and evaluation in engineering applications.
基金This work has been funded by the National Key Research and Development Project(2017YFC0307605,2017YFC0307702),National Science and Technology Major Project of China(No.2017ZX05008-008)and State Key Project(2016ZX05051004).The authors also thank Sinopec Oilfield Service Jianghan Corporation for providing the data and support and express their gratitude to the reviewers for their constructive comments.
文摘Microseismic phase detection and arrival picking are critical steps in the processing of hydraulic fracturing microseismic monitoring data.As the signal-to-noise ratios of P-and S-waves diff er because of the infl uence of focal mechanisms,geometry,and background noise,diffi culties are introduced in the eff ective discrimination of seismic phases and the accurate acquisition of arrivals in conventional processing methods.In this paper,we propose a method for identifying microseismic phase arrival time by comprehensively analyzing the variation of moveout curves and combining the intra-event waveform similarities with the energy ratio of multitrace signals.First,a curve-fi tting formula is constructed with perforation arrivals,and event detection is achieved by adopting an energy-weighted similarity coeffi cient that seeks plausible fi tting curves with a sliding time window in continuous microseismic recordings.Then,the P-and S-waves are separated by the fitting parameters.The known arrival time trend of the microseismic phase is employed to calculate residual time corrections.Finally,the accurate arrival results of the microseismic phases can be obtained by picking the arrivals of stacked traces.The reliability and eff ectiveness of the proposed method for microseismic phase detection and arrival picking were determined through tests using field data.Arrival results indicate that the proposed method can improve accuracy compared with the traditional energy ratio method.
基金supported by the National Major Scientific and Technological Special Project during the 13th Five-year Plan Period(No.2016ZX05045003-005)
文摘Accurately detecting the arrival time of a channel wave in a coal seam is very important for in-seam seismic data processing. The arrival time greatly affects the accuracy of the channel wave inversion and the computed tomography （CT） result. However, because the signal-to-noise ratio of in-seam seismic data is reduced by the long wavelength and strong frequency dispersion, accurately timing the arrival of channel waves is extremely difficult. For this purpose, we propose a method that automatically picks up the arrival time of channel waves based on multi-channel constraints. We first estimate the Jaccard similarity coefficient of two ray paths, then apply it as a weight coefficient for stacking the multi- channel dispersion spectra. The reasonableness and effectiveness of the proposed method is verified in an actual data application. Most importantly, the method increases the degree of automation and the pickup precision of the channel-wave arrival time.
文摘基于微机电系统(Micro-Electro-Mechanical System,MEMS)技术研制的MEMS强震仪具有易集成、维护成本低和低功耗等优点,在地震监测领域应用广泛.然而,MEMS强震仪集成的软、硬件资源有限,并且受仪器自身噪声等因素干扰较大,地震信号测量结果质量较低,对嵌入算法要求更高.针对这一问题,本文提出一种更适用于MEMS强震仪的改进长短时窗均值比(Short Term Average/Long Term Average,STA/LTA)算法.首先,通过构建抗干扰(Anti-interference,AR)特征函数抑制基线漂移和低频噪声的干扰,提高STA/LTA算法拾取地震事件的抗干扰能力;其次,提出采用“延时长窗”的方式,提高STA/LTA算法的计算效率和拾取精度,减少STA/LTA算法对MEMS集成资源的占用;最后,结合时窗位置进一步探究不同时窗大小对STA/LTA算法拾取效率的影响.实际地震资料处理结果表明,本文提出的改进STA/LTA算法计算效率更高,实时性和抗干扰能力更强,更适用于集成资源有限的MEMS强震仪.
基金funded by the Deep Earth Probe and Mineral Resources Exploration-National Science and Technology Major Project(Grant No.2024ZD1002503)。
文摘Passive microseismic monitoring(PMM)serves as a fundamental technology for assessing hydraulic fracturing(HF)effectiveness,with a key focus on accurate and efficient phase detection/arrival picking and source location.In PMM data processing,the data-driven paradigm(deep learning based)outperforms the model-driven paradigm in characteristic extraction but lacks quality control and uncertainty quantification.Monte Carlo Dropout,a Bayesian uncertainty quantification technique,performs stochastic neuron deactivation through multiple forward propagation samplings.Therefore,this study proposes a deep learning neural network incorporating uncertainty quantification with manual quality control integration,establishing an optimized workflow spanning automated phase detection to robust source location.The methodology implementation comprises two principal components:(1)The MDNet employing Monte Carlo Dropout strategy enabling simultaneous phase detection/arrival picking and unce rtainty estimation;(2)an integrated hybrid-driven workflow with a traveltime-based inve rsion method for source location.Validation with field data demonstrates that MD-Net achieves superior performance under low signal-to-noise ratio conditions,maintaining detection accuracy exceeding 99%for both P-and S-waves.The phase arrival picking precision shows significant improvement,with a 40%reduction in standard deviation compared to the baseline model(P-S time difference decreasing from12.0 ms to 7.1 ms),while providing quantifiable uncertainty metrics for manual calibration.Source location results further reveal that our hybrid-driven workflow produces more physically plausible event distributions,with 100%of microseismic eve nts clustering along the primary fracture expanding direction.This performance surpasses traditional cross-correlation methods and single/multi-trace data-driven me thods in spatial rationality.This study establishes an inte rpretable,high-pre cision automated framework for HF-PMM applications,demonstrating potential for extension to diverse geological settings and monitoring configurations.
