Rockburst is a common dynamic geological hazard,frequently occurring in underground engineering(e.g.,TBM tunnelling and deep mining).In order to achieve rockburst monitoring and warning,the microseismic moni-toring te...Rockburst is a common dynamic geological hazard,frequently occurring in underground engineering(e.g.,TBM tunnelling and deep mining).In order to achieve rockburst monitoring and warning,the microseismic moni-toring technique has been widely used in the field.However,the microseismic source location has always been a challenge,playing a vital role in the precise prevention and control of rockburst.To this end,this study proposes a novel microseismic source location model that considers the anisotropy of P-wave velocity.On the one hand,it assigns a unique P-wave velocity to each propagation path,abandoning the assumption of a homogeneous ve-locity field.On the other hand,it treats the P-wave velocity as a co-inversion parameter along with the source location,avoiding the predetermination of P-wave velocity.To solve this model,three various metaheuristic multi-objective optimization algorithms are integrated with it,including the whale optimization algorithm,the butterfly optimization algorithm,and the sparrow search algorithm.To demonstrate the advantages of the model in terms of localization accuracy,localization efficiency,and solution stability,four blasting cases are collected from a water diversion tunnel project in Xinjiang,China.Finally,the effect of the number of involved sensors on the microseismic source location is discussed.展开更多
Through a case analysis,this study examines the spatiotemporal evolution of microseismic(MS)events,energy characteristics,volumetric features,and fracture network development in surface well hydraulic fracturing.A tot...Through a case analysis,this study examines the spatiotemporal evolution of microseismic(MS)events,energy characteristics,volumetric features,and fracture network development in surface well hydraulic fracturing.A total of 349 MS events were analyzed across different fracturing sections,revealing significant heterogeneity in fracture propagation.Energy scanning results showed that cumulative energy values ranged from 240 to 1060 J across the sections,indicating notable differences.Stimulated reservoir volume(SRV)analysis demonstrated well-developed fracture networks in certain sections,with a total SRV exceeding 1540000 m^(3).The hydraulic fracture network analysis revealed that during the midfracturing stage,the density and spatial extent of MS events significantly increased,indicating rapid fracture propagation and the formation of complex networks.In the later stage,the number of secondary fractures near fracture edges decreased,and the fracture network stabilized.By comparing the branching index,fracture length,width,height,and SRV values across different fracturing sections,Sections No.1 and No.8 showed the best performance,with high MS event densities,extensive fracture networks,and significant energy release.However,Sections No.4 and No.5 exhibited sparse MS activity and poor fracture connectivity,indicating suboptimal stimulation effectiveness.展开更多
This is a continuation of the article“Ground Monitoring of Microseismic Based on Low Signal-to-Noise Ratio”,and a further summary and reflection after investigating the current situation of microseismic monitoring.I...This is a continuation of the article“Ground Monitoring of Microseismic Based on Low Signal-to-Noise Ratio”,and a further summary and reflection after investigating the current situation of microseismic monitoring.It is difficult to provide necessary and sufficient conditions to test the reliability of microseismic monitoring.Often,a few hundred meters away,the microseismic signal emitted by a hypocenter is submerged in noise,and the traditional location is invalid;Inversion for microseismic released energy distribution using data migration and stacking is in principle not unique.However,based on microseismic monitoring characteristics,forward and reverse simulations and numerous experiments,many necessary conditions can be proposed to ensure reliable monitoring with high probability.VS(Vector Scanning)ground monitoring for microseismic proposes eight necessary conditions for testing the reliability,so that VS finds the fracturing-induced effective communication seam with the characteristics of shear zones under the control of tectonic stress fields,in line with the laws of seismic and geological observations,as well as the features related to some special production data.VS uses data migration and stacking suitable for low signal-to-noise ratio and shear mechanism,and the joint inversion for correction of both traditional relocations and velocity model,can greatly improve monitoring distance and quality,complete microseismic measurement methods,and broaden applicable fields,such as:(1)VS can be a cost-effective,ground-based,routine monitoring method;(2)The BPM(Borehole Proximity Monitoring)is high cost but close to the hypocenters;It can be the best method for scientific research,but its seismic network should be improved,and the joint inversion and data stacking could be used to improve the monitoring distance and quality;(3)The early warning of mine safety can change the current monitoring of strong microseismic(or accidents have been happened)to the real microseismic level;and(4)The seismic precursor monitoring of large earthquakes can be expanded from small earthquakes to microseismic.These will establish a solid foundation and complete seismic measurements for microseismology.展开更多
This paper proposed a moment tensor regression prediction technology based on ResNet for microseismic events.Taking the great advantages of deep networks in classification and regression tasks,it can realize the great...This paper proposed a moment tensor regression prediction technology based on ResNet for microseismic events.Taking the great advantages of deep networks in classification and regression tasks,it can realize the great potential of fast and accurate inversion of microseismic moment tensors after the network trained.This ResNet-based moment tensor prediction technology,whose input is raw recordings,does not require the extraction of data features in advance.First,we tested the network using synthetic data and performed a quantitative assessment of the errors.The results demonstrate that the network exhibits high accuracy and efficiency during the prediction phase.Next,we tested the network using real microseismic data and compared the results with those from traditional inversion methods.The error in the results was relatively small compared to traditional methods.However,the network operates more efficiently without requiring manual intervention,making it highly valuable for near-real-time monitoring applications.展开更多
Microseismic monitoring is essential for understanding subsurface dynamics and optimizing oil and gas pro-duction.However,traditional methods for the automatic detection of microseismic events rely heavily on characte...Microseismic monitoring is essential for understanding subsurface dynamics and optimizing oil and gas pro-duction.However,traditional methods for the automatic detection of microseismic events rely heavily on characteristic functions and human intervention,often resulting in suboptimal performance when dealing with complex and noisy data.In this study,we propose a novel approach that leverages deep learning frame to extract multiscale features from microseismic data using a TransUNet neural network.Our model integrates the ad-vantages of Transformer and UNet architectures to achieve high accuracy in multivariate image segmentation and precise picking of P-wave and S-wave first arrivals simultaneously.We validate our approach using both synthetic and field microseismic datasets recorded from gas storage monitoring and roof fracturing in a coal seam.The robustness of the proposed method has been verified in the testing of synthetic data with various levels of Gaussian and real background noises extracted from field data.The comparisons of the proposed method with UNet and SwinUNet in terms of the model architecture and classification performance demonstrate the Tran-sUNet achieves the optimal balance in its architecture and inference speed.With relatively low inference time and network complexity,it operates effectively in high-precision microseismic phase pickings.This advancement holds significant promise for enhancing microseismic monitoring technology in hydraulic fracturing and reser-voir monitoring applications.展开更多
The collapse of rock masses in fault-developed zones poses significant safety challenges during the excavation of high-stress underground caverns. This study investigates the spatiotemporal evolution of the collapse m...The collapse of rock masses in fault-developed zones poses significant safety challenges during the excavation of high-stress underground caverns. This study investigates the spatiotemporal evolution of the collapse mechanisms of the cavern in the Yebatan Hydropower Station through using microseismic (MS) monitoring and displacement measurements. We developed a multi-parameter deformation early warning model that integrates three critical indicators: deformation rate, rate increment, and tangential angle of the deformation time curve. The results of the early warning model show a significant and abrupt increase in the deformation of the rock mass during the collapse process. The safety and stability of the local cavern in the face of excavation-induced disturbances are meticulously assessed utilizing MS data. Spatiotemporal analysis of the MS monitoring indicates a high frequency of MS events during the blasting phase, with a notable clustering of these events in the vicinity of the fault. These research results provide a valuable reference for risk warnings and stability assessments in the fault development zones of analogous caverns.展开更多
With the continuous expansion of deep underground engineering and the growing demand for safety monitoring,microseismic monitoring has become a core method for early warning of rock mass fracture and engineering stabi...With the continuous expansion of deep underground engineering and the growing demand for safety monitoring,microseismic monitoring has become a core method for early warning of rock mass fracture and engineering stability assessment.To address problems in existing methods,such as low data processing efficiency and poor phase recognition accuracy under low signal-to-noise ratio(SNR)conditions in complex geological environments,this study proposes an intelligent phase picking model based on ResUNet.The model integrates the residual learning mechanism of ResNet with the multi-scale feature extraction capability of UNet,effectively mitigating the vanishing gradient problem in deep networks.It also achieves cross-layer fusion of shallow detail features and deep semantic features through skip connections in the encoder-decoder structure.Compared with traditional short-time average/long-time average(STA/LTA)algorithms and advanced neural network models such as PhaseNet and EQTransformer,ResUNet shows superior performance in picking P-and S-wave phases.The model was trained on 400000 labeled microseismic signals from the Stanford earthquake dataset(STEAD)and was successfully applied to the Shizhuyuan polymetallic mine in Hunan Province,China.The results demonstrate that ResUNet achieves high picking accuracy and robustness in complex geological conditions,offering reliable technical support for early warning of disasters such as rockburst in deep underground engineering.展开更多
Microseismic (MS) source location plays an important role in MS monitoring. This paper proposes a MS source location method based on particle swarm optimization (PSO) and multi-sensor arrays, where a free weight joint...Microseismic (MS) source location plays an important role in MS monitoring. This paper proposes a MS source location method based on particle swarm optimization (PSO) and multi-sensor arrays, where a free weight joints the P-wave first arrival data. This method adaptively adjusts the preference for “superior” arrays and leverages “inferior” arrays to escape local optima, thereby improving the location accuracy. The effectiveness and stability of this method were validated through synthetic tests, pencil-lead break (PLB) experiments, and mining engineering applications. Specifically, for synthetic tests with 1 μs Gaussian noise and 100 μs large noise in rock samples, the location error of the multi-sensor arrays jointed location method is only 0.30 cm, which improves location accuracy by 97.51% compared to that using a single sensor array. The average location error of PLB events on three surfaces of a rock sample is reduced by 48.95%, 26.40%, and 55.84%, respectively. For mine blast event tests, the average location error of the dual sensor arrays jointed method is 62.74 m, 54.32% and 14.29% lower than that using only sensor arrays 1 and 2, respectively. In summary, the proposed multi-sensor arrays jointed location method demonstrates good noise resistance, stability, and accuracy, providing a compelling new solution for MS location in relevant mining scenarios.展开更多
A case study of seismic interferometry applied to a small microseismic monitoring network is here presented.The main objectives of this study are(i)to quantify the lateral variability of shear-wave ve-locities in the ...A case study of seismic interferometry applied to a small microseismic monitoring network is here presented.The main objectives of this study are(i)to quantify the lateral variability of shear-wave ve-locities in the studied area,and(ii)to investigate the bias produced by noise directionality and non-stationarity in the velocity estimate.Despite the limited number of stations and the short-period char-acter of the seismic sensors,the empirical Green's functions were retrieved for all station pairs using two years of passive data.Both group and phase velocities were derived,the former using the widespread frequency-time analysis,the latter through the analysis of the real part of the cross-spectra.The main advantage of combining these two methods is a more accurate identification of higher modes,resulting in a reduction of ambiguity during picking and data interpretation.Surface wave tomography was run to obtain the spatial distribution of group and phase velocities for the same wavelengths.The low standard deviation of the results suggests that the sparse character of the network does not limit the applicability of the method,for this specific case.The obtained maps highlight the presence of a lower velocity area that extends from the centre of the network towards southeast.Group and phase velocity dispersion curves have been jointly inverted to retrieve as many shear-wave velocity profiles as selected station pairs.While the average model can be used for a more accurate location of the local natural seismicity,the associated standard deviations give us an indication of the lateral heterogeneity of seismic velocities as a function of depth.Finally,the same velocity analysis was repeated for different time windows in order to quantify the error associated to variations in the noise field.Errors as large as 4%have been found,related to the unfavorable orientation of the receiver pairs with respect to strongly directional noise sources,and to the very short time widows.It was shown that using a one-year time window these errors arereduced to 0.3%.展开更多
Deep coal-energy mining frequently results in microseismic(MS)events,which may be a precursor to the risk of rockbursts and pose risks to human safety and infrastructure.Therefore,quantitatively predicting the time,en...Deep coal-energy mining frequently results in microseismic(MS)events,which may be a precursor to the risk of rockbursts and pose risks to human safety and infrastructure.Therefore,quantitatively predicting the time,energy,and location(TEL)of future MS events is crucial for understanding and preventing potential catastrophic events.In this study,we introduced the application of spatiotemporal graph convolutional networks(STGCN)to predict the TEL of MS events induced by deep coal-energy mining.Notably,this was the first application of graph convolution networks(GCNs)in the spatiotemporal prediction of MS events.The adjacency matrices of the sensor networks were determined based on the distance between MS sensors,the sensor network graphs we constructed,and GCN was employed to extract the spatiotemporal details of the graphs.The model is simple and versatile.By testing the model with on-site MS monitoring data,our results demonstrated promising efficacy in predicting the TEL of MS events,with the cosine similarity(C)above 0.90 and the mean relative error(MRE)below 0.08.This is critical to improving the safety and operational efficiency of deep coal-energy mining.展开更多
Dynamic stress adjustment in deep-buried high geostress hard rock tunnels frequently triggers catastrophic failures such as rockbursts and collapses.While a comprehensive understanding of this process is critical for ...Dynamic stress adjustment in deep-buried high geostress hard rock tunnels frequently triggers catastrophic failures such as rockbursts and collapses.While a comprehensive understanding of this process is critical for evaluating surrounding rock stability,its dynamic evolution are often overlooked in engineering practice.This study systematically summarizes a novel classification framework for stress adjustment types—stabilizing(two-zoned),shallow failure(three-zoned),and deep failure(four-zoned)—characterized by distinct stress adjustment stages.A dynamic interpretation technology system is developed based on microseismic monitoring,integrating key microseismic parameters(energy index EI,apparent stressσa,microseismic activity S),seismic source parameter space clustering,and microseismic paths.This approach enables precise identification of evolutionary stages,stress adjustment types,and failure precursors,thereby elucidating the intrinsic linkage between geomechanical processes(stress redistribution)and failure risks.The study establishes criteria and procedures for identifying stress adjustment types and their associated failure risks,which were successfully applied in the Grand Canyon Tunnel of the E-han Highway to detect 50 instances of disaster risks.The findings offer invaluable insights into understanding the evolution process of stress adjustment and pinpointing the disaster risks linked to hard rock in comparable high geostress tunnels.展开更多
Accurate and rapid determination of source locations is of great significance for surface microseismic monitoring.Traditional methods,such as diffraction stacking,are time-consuming and challenging for real-time monit...Accurate and rapid determination of source locations is of great significance for surface microseismic monitoring.Traditional methods,such as diffraction stacking,are time-consuming and challenging for real-time monitoring.In this study,we propose an approach to locate microseismic events using a deep learning algorithm with surface data.A fully convolutional network is designed to predict source locations.The input data is the waveform of a microseismic event,and the output consists of three 1D Gaussian distributions representing the probability distribution of the source location in the x,y,and z dimensions.The theoretical dataset is generated to train the model,and several data augmentation methods are applied to reduce discrepancies between the theoretical and field data.After applying the trained model to field data,the results demonstrate that our method is fast and achieves comparable location accuracy to the traditional diffraction stacking location method,making it promising for real-time microseismic monitoring.展开更多
In this study, we use the Bohai Sea area as an example to investigate the characteristics of secondary microseisms and their impact on seismic noise based on the temporal frequency spectral analysis of observation dat...In this study, we use the Bohai Sea area as an example to investigate the characteristics of secondary microseisms and their impact on seismic noise based on the temporal frequency spectral analysis of observation data from 33 broadband seismic stations during strong gust periods, and new perspectives are proposed on the generation mechanisms of secondary microseisms. The results show that short-period double- frequency (SPDF) and long-period double-frequency (LPDF) microseisms exhibit significant alternating trends of strengthening and weakening in the northwest area of the Bohai Sea. SPDF microseisms are generated by irregular wind waves during strong off shore wind periods, with a broad frequency band distributed in the range of 0.2-1 Hz;LPDF microseisms are generated by regular swells during periods of sea wind weakening, with a narrow frequency band concentrated between 0.15 and 0.3 Hz. In terms of temporal dimensions, as the sea wind weakens, the energy of SPDF microseisms weakens, and the dominant frequencies increase, whereas the energy of LPDF microseisms strengthens and the dominant frequencies decrease, which is consistent with the process of the decay of wind waves and the growth of swells. In terms of spatial dimensions, as the microseisms propagate inland areas, the advantageous frequency band and energy of SPDF microseisms are reduced and significantly attenuated, respectively, whereas LPDF microseisms show no significant changes. And during the propagation process in high-elevation areas, LPDF microseisms exhibit a certain site amplifi cation eff ect when the energy is strong. The results provide important supplements to the basic theory of secondary microseisms, preliminarily reveal the relationship between the atmosphere, ocean, and seismic noise, and provide important theoretical references for conducting geological and oceanographic research based on the characteristics of secondary microseisms.展开更多
The rock mass failure induced by deep mining exhibits pronounced spatial heterogeneity and diverse mechanisms,with its microseismic responses serving as effective indicators of regional failure evolution and instabili...The rock mass failure induced by deep mining exhibits pronounced spatial heterogeneity and diverse mechanisms,with its microseismic responses serving as effective indicators of regional failure evolution and instability mechanisms.Focusing on the Level VI stope sublayers in the Jinchuan#2 mining area,this study constructs a 24-parameter index system encompassing time-domain features,frequency-domain features,and multifractal characteristics.Through manifold learning,clustering analysis,and hybrid feature selection,15 key indicators were extracted to construct a classification framework for failure responses.Integrated with focal mechanism inversion and numerical simulation,the failure patterns and corresponding instability mechanisms across different structural zones were further identified.The results reveal that multiscale microseismic characteristics exhibit clear regional similarities.Based on the morphological features of radar plots derived from the 15 indicators,acoustic responses were classified into four typical types,each reflecting distinct local failure mechanisms,stress conditions,and plastic zone evolution.Moreover,considering dominant instability factors and rupture modes,four representative rock mass instability models were proposed for typical failure zones within the stope.These findings provide theoretical guidance and methodological support for hazard prediction,structural optimization,and disturbance control in deep metal mining areas.展开更多
In deeply buried tunneling projects,geological conditions are often complex and varied.Microseismic monitoring systems are extensively deployed to enhance construction safety.However,when the current geological condit...In deeply buried tunneling projects,geological conditions are often complex and varied.Microseismic monitoring systems are extensively deployed to enhance construction safety.However,when the current geological conditions differ from those present during the signal collection for model training,recognition accuracy tends to decline significantly.Therefore,improving the applicability and stability of microseismic waveform recognition models across varying geological conditions has emerged as a critical challenge.To address this issue,we first analyze the impact of lithological changes and the development of structural planes on the features of microseismic waveforms.Subsequently,we propose a category-domain-aligned transfer learning method that enables the transfer of recognition capabilities across geological conditions by facilitating similar feature extraction and the recognition of cross-geological fracture waveforms.In this model,feature separation modeling enhances the extraction of category features of waveforms under different geological conditions.A deep transfer learning mechanism distinguishes between unique and common features,allowing for the capture of essential features necessary for model parameter updates.Through comparative experiments and feature distribution alignment and visualization,we demonstrate that the accuracy of microseismic waveform recognition across geological conditions achieves 90%.Additionally,the performance of our method is validated using microseismic signals collected from different sections of the construction site.展开更多
Monitoring sensors in complex engineering environments often record abnormal data,leading to significant positioning errors.To reduce the influence of abnormal arrival times,we introduce an innovative,outlier-robust l...Monitoring sensors in complex engineering environments often record abnormal data,leading to significant positioning errors.To reduce the influence of abnormal arrival times,we introduce an innovative,outlier-robust localization method that integrates kernel density estimation(KDE)with damping linear correction to enhance the precision of microseismic/acoustic emission(MS/AE)source positioning.Our approach systematically addresses abnormal arrival times through a three-step process:initial location by 4-arrival combinations,elimination of outliers based on three-dimensional KDE,and refinement using a linear correction with an adaptive damping factor.We validate our method through lead-breaking experiments,demonstrating over a 23%improvement in positioning accuracy with a maximum error of 9.12 mm(relative error of 15.80%)—outperforming 4 existing methods.Simulations under various system errors,outlier scales,and ratios substantiate our method’s superior performance.Field blasting experiments also confirm the practical applicability,with an average positioning error of 11.71 m(relative error of 7.59%),compared to 23.56,66.09,16.95,and 28.52 m for other methods.This research is significant as it enhances the robustness of MS/AE source localization when confronted with data anomalies.It also provides a practical solution for real-world engineering and safety monitoring applications.展开更多
Microseism,acoustic emission and electromagnetic radiation(M-A-E)data are usually used for predicting rockburst hazards.However,it is a great challenge to realize the prediction of M-A-E data.In this study,with the ai...Microseism,acoustic emission and electromagnetic radiation(M-A-E)data are usually used for predicting rockburst hazards.However,it is a great challenge to realize the prediction of M-A-E data.In this study,with the aid of a deep learning algorithm,a new method for the prediction of M-A-E data is proposed.In this method,an M-A-E data prediction model is built based on a variety of neural networks after analyzing numerous M-A-E data,and then the M-A-E data can be predicted.The predicted results are highly correlated with the real data collected in the field.Through field verification,the deep learning-based prediction method of M-A-E data provides quantitative prediction data for rockburst monitoring.展开更多
Microseismic source location is crucial for the early warning of rockburst risks.However,the conventional methods face challenges in terms of the microseismic wave velocity and arrival time accuracy.Intelligent techni...Microseismic source location is crucial for the early warning of rockburst risks.However,the conventional methods face challenges in terms of the microseismic wave velocity and arrival time accuracy.Intelligent techniques,such as the full convolutional neural network(FCNN),can capture spatial information but struggle with complex microseismic sequence.Combining the FCNN with the long shortterm memory(LSTM)network enables better time-series signal classification by integrating multiscale information and is therefore suitable for waveform location.The LSTM-FCNN model does not require extensive data preprocessing and it simplifies the microseismic source location through feature extraction.In this study,we utilized the LSTM-FCNN as a regression learning model to locate the seismic focus.Initially,the method of short-time-average/long-time-average(STA/LTA)arrival time picking was employed to augment spatiotemporal information.Subsequently,oversampling the on-site data was performed to address the issue of data imbalance,and finally,the performance of LSTM-FCNN was tested.Meanwhile,we compared the LSTM-FCNN model with previous deep-learning models.Our results demonstrated remarkable location capabilities with a mean absolute error(MAE)of only 7.16 m.The model can realize swift training and high accuracy,thereby significantly improving risk warning of rockbursts.展开更多
Rock fracture mechanisms can be inferred from moment tensors(MT)inverted from microseismic events.However,MT can only be inverted for events whose waveforms are acquired across a network of sensors.This is limiting fo...Rock fracture mechanisms can be inferred from moment tensors(MT)inverted from microseismic events.However,MT can only be inverted for events whose waveforms are acquired across a network of sensors.This is limiting for underground mines where the microseismic stations often lack azimuthal coverage.Thus,there is a need for a method to invert fracture mechanisms using waveforms acquired by a sparse microseismic network.Here,we present a novel,multi-scale framework to classify whether a rock crack contracts or dilates based on a single waveform.The framework consists of a deep learning model that is initially trained on 2400000+manually labelled field-scale seismic and microseismic waveforms acquired across 692 stations.Transfer learning is then applied to fine-tune the model on 300000+MT-labelled labscale acoustic emission waveforms from 39 individual experiments instrumented with different sensor layouts,loading,and rock types in training.The optimal model achieves over 86%F-score on unseen waveforms at both the lab-and field-scale.This model outperforms existing empirical methods in classification of rock fracture mechanisms monitored by a sparse microseismic network.This facilitates rapid assessment of,and early warning against,various rock engineering hazard such as induced earthquakes and rock bursts.展开更多
The denoising of microseismic signals is a prerequisite for subsequent analysis and research.In this research,a new microseismic signal denoising algorithm called the Black Widow Optimization Algorithm(BWOA)optimized ...The denoising of microseismic signals is a prerequisite for subsequent analysis and research.In this research,a new microseismic signal denoising algorithm called the Black Widow Optimization Algorithm(BWOA)optimized VariationalMode Decomposition(VMD)jointWavelet Threshold Denoising(WTD)algorithm(BVW)is proposed.The BVW algorithm integrates VMD and WTD,both of which are optimized by BWOA.Specifically,this algorithm utilizes VMD to decompose the microseismic signal to be denoised into several Band-Limited IntrinsicMode Functions(BLIMFs).Subsequently,these BLIMFs whose correlation coefficients with the microseismic signal to be denoised are higher than a threshold are selected as the effective mode functions,and the effective mode functions are denoised using WTD to filter out the residual low-and intermediate-frequency noise.Finally,the denoised microseismic signal is obtained through reconstruction.The ideal values of VMD parameters and WTD parameters are acquired by searching with BWOA to achieve the best VMD decomposition performance and solve the problem of relying on experience and requiring a large workload in the application of the WTD algorithm.The outcomes of simulated experiments indicate that this algorithm is capable of achieving good denoising performance under noise of different intensities,and the denoising performance is significantly better than the commonly used VMD and Empirical Mode Decomposition(EMD)algorithms.The BVW algorithm is more efficient in filtering noise,the waveform after denoising is smoother,the amplitude of the waveform is the closest to the original signal,and the signal-to-noise ratio(SNR)and the root mean square error after denoising are more satisfying.The case based on Fankou Lead-Zinc Mine shows that for microseismic signals with different intensities of noise monitored on-site,compared with VMD and EMD,the BVW algorithm ismore efficient in filtering noise,and the SNR after denoising is higher.展开更多
基金supported by the National Natural Science Founda-tion of China under Grant Nos.42472351,42177140,52404127,and 42207235the Natural Science Foundation of Hubei Province under Grant No.2024AFD359+1 种基金the Young Elite Scientist Sponsorship Program by CAST under Grant No.YESS20230742the China Postdoctoral Science Foundation Program under Grant No.2024T170684.
文摘Rockburst is a common dynamic geological hazard,frequently occurring in underground engineering(e.g.,TBM tunnelling and deep mining).In order to achieve rockburst monitoring and warning,the microseismic moni-toring technique has been widely used in the field.However,the microseismic source location has always been a challenge,playing a vital role in the precise prevention and control of rockburst.To this end,this study proposes a novel microseismic source location model that considers the anisotropy of P-wave velocity.On the one hand,it assigns a unique P-wave velocity to each propagation path,abandoning the assumption of a homogeneous ve-locity field.On the other hand,it treats the P-wave velocity as a co-inversion parameter along with the source location,avoiding the predetermination of P-wave velocity.To solve this model,three various metaheuristic multi-objective optimization algorithms are integrated with it,including the whale optimization algorithm,the butterfly optimization algorithm,and the sparrow search algorithm.To demonstrate the advantages of the model in terms of localization accuracy,localization efficiency,and solution stability,four blasting cases are collected from a water diversion tunnel project in Xinjiang,China.Finally,the effect of the number of involved sensors on the microseismic source location is discussed.
基金supported by Yunlong Lake Laboratory of Deep Underground Science and Engineering Project(No.104024008)the National Natural Science Foundation of China(Nos.52274241 and 52474261)the Natural Science Foundation of Jiangsu Province(No.BK20240207).
文摘Through a case analysis,this study examines the spatiotemporal evolution of microseismic(MS)events,energy characteristics,volumetric features,and fracture network development in surface well hydraulic fracturing.A total of 349 MS events were analyzed across different fracturing sections,revealing significant heterogeneity in fracture propagation.Energy scanning results showed that cumulative energy values ranged from 240 to 1060 J across the sections,indicating notable differences.Stimulated reservoir volume(SRV)analysis demonstrated well-developed fracture networks in certain sections,with a total SRV exceeding 1540000 m^(3).The hydraulic fracture network analysis revealed that during the midfracturing stage,the density and spatial extent of MS events significantly increased,indicating rapid fracture propagation and the formation of complex networks.In the later stage,the number of secondary fractures near fracture edges decreased,and the fracture network stabilized.By comparing the branching index,fracture length,width,height,and SRV values across different fracturing sections,Sections No.1 and No.8 showed the best performance,with high MS event densities,extensive fracture networks,and significant energy release.However,Sections No.4 and No.5 exhibited sparse MS activity and poor fracture connectivity,indicating suboptimal stimulation effectiveness.
文摘This is a continuation of the article“Ground Monitoring of Microseismic Based on Low Signal-to-Noise Ratio”,and a further summary and reflection after investigating the current situation of microseismic monitoring.It is difficult to provide necessary and sufficient conditions to test the reliability of microseismic monitoring.Often,a few hundred meters away,the microseismic signal emitted by a hypocenter is submerged in noise,and the traditional location is invalid;Inversion for microseismic released energy distribution using data migration and stacking is in principle not unique.However,based on microseismic monitoring characteristics,forward and reverse simulations and numerous experiments,many necessary conditions can be proposed to ensure reliable monitoring with high probability.VS(Vector Scanning)ground monitoring for microseismic proposes eight necessary conditions for testing the reliability,so that VS finds the fracturing-induced effective communication seam with the characteristics of shear zones under the control of tectonic stress fields,in line with the laws of seismic and geological observations,as well as the features related to some special production data.VS uses data migration and stacking suitable for low signal-to-noise ratio and shear mechanism,and the joint inversion for correction of both traditional relocations and velocity model,can greatly improve monitoring distance and quality,complete microseismic measurement methods,and broaden applicable fields,such as:(1)VS can be a cost-effective,ground-based,routine monitoring method;(2)The BPM(Borehole Proximity Monitoring)is high cost but close to the hypocenters;It can be the best method for scientific research,but its seismic network should be improved,and the joint inversion and data stacking could be used to improve the monitoring distance and quality;(3)The early warning of mine safety can change the current monitoring of strong microseismic(or accidents have been happened)to the real microseismic level;and(4)The seismic precursor monitoring of large earthquakes can be expanded from small earthquakes to microseismic.These will establish a solid foundation and complete seismic measurements for microseismology.
基金supported by the National Natural Science dation Foun-of China(Grant Number 42272204)Key Laboratory of Coal sources Re-Exploration and Comprehensive Utilization,Ministry of Natural Resources,Canada(SMDZ-KF2024-4)+1 种基金the Fundamental Research Funds for the Central Universities,China(Grant No.2024JCCXDC06)supported in part by open fund project of State Key Laboratory for Fine Exploration and Intelligent Development of Coal Research(SKLCRSM23KFA04)。
文摘This paper proposed a moment tensor regression prediction technology based on ResNet for microseismic events.Taking the great advantages of deep networks in classification and regression tasks,it can realize the great potential of fast and accurate inversion of microseismic moment tensors after the network trained.This ResNet-based moment tensor prediction technology,whose input is raw recordings,does not require the extraction of data features in advance.First,we tested the network using synthetic data and performed a quantitative assessment of the errors.The results demonstrate that the network exhibits high accuracy and efficiency during the prediction phase.Next,we tested the network using real microseismic data and compared the results with those from traditional inversion methods.The error in the results was relatively small compared to traditional methods.However,the network operates more efficiently without requiring manual intervention,making it highly valuable for near-real-time monitoring applications.
基金supported by a National Natural Science Foundation of China(Grant number 41974150 and 42174158)Natural Science Basic Research Program of Shaanxi(2023-JC-YB-220).
文摘Microseismic monitoring is essential for understanding subsurface dynamics and optimizing oil and gas pro-duction.However,traditional methods for the automatic detection of microseismic events rely heavily on characteristic functions and human intervention,often resulting in suboptimal performance when dealing with complex and noisy data.In this study,we propose a novel approach that leverages deep learning frame to extract multiscale features from microseismic data using a TransUNet neural network.Our model integrates the ad-vantages of Transformer and UNet architectures to achieve high accuracy in multivariate image segmentation and precise picking of P-wave and S-wave first arrivals simultaneously.We validate our approach using both synthetic and field microseismic datasets recorded from gas storage monitoring and roof fracturing in a coal seam.The robustness of the proposed method has been verified in the testing of synthetic data with various levels of Gaussian and real background noises extracted from field data.The comparisons of the proposed method with UNet and SwinUNet in terms of the model architecture and classification performance demonstrate the Tran-sUNet achieves the optimal balance in its architecture and inference speed.With relatively low inference time and network complexity,it operates effectively in high-precision microseismic phase pickings.This advancement holds significant promise for enhancing microseismic monitoring technology in hydraulic fracturing and reser-voir monitoring applications.
基金Projects(52209132, 52309156) supported by the National Natural Science Foundation of ChinaProject(BK20251905) supported by the Natural Science Foundation of Jiangsu Province,China+2 种基金Project(252102320037) supported by the Henan Province Science and Technology Research,ChinaProject(CKWV20231173/KY) supported by the CRSRI Open Research Program,ChinaProject(2023KSD15) supported by the Open Research Fund of Hubei Provincial Key Laboratory of Construction and Management in Hydropower Engineering,China。
文摘The collapse of rock masses in fault-developed zones poses significant safety challenges during the excavation of high-stress underground caverns. This study investigates the spatiotemporal evolution of the collapse mechanisms of the cavern in the Yebatan Hydropower Station through using microseismic (MS) monitoring and displacement measurements. We developed a multi-parameter deformation early warning model that integrates three critical indicators: deformation rate, rate increment, and tangential angle of the deformation time curve. The results of the early warning model show a significant and abrupt increase in the deformation of the rock mass during the collapse process. The safety and stability of the local cavern in the face of excavation-induced disturbances are meticulously assessed utilizing MS data. Spatiotemporal analysis of the MS monitoring indicates a high frequency of MS events during the blasting phase, with a notable clustering of these events in the vicinity of the fault. These research results provide a valuable reference for risk warnings and stability assessments in the fault development zones of analogous caverns.
基金Project(2022YFC2905100)supported by the National Key Research and Development Program of ChinaProject(52174098)supported by the National Natural Science Foundation of China。
文摘With the continuous expansion of deep underground engineering and the growing demand for safety monitoring,microseismic monitoring has become a core method for early warning of rock mass fracture and engineering stability assessment.To address problems in existing methods,such as low data processing efficiency and poor phase recognition accuracy under low signal-to-noise ratio(SNR)conditions in complex geological environments,this study proposes an intelligent phase picking model based on ResUNet.The model integrates the residual learning mechanism of ResNet with the multi-scale feature extraction capability of UNet,effectively mitigating the vanishing gradient problem in deep networks.It also achieves cross-layer fusion of shallow detail features and deep semantic features through skip connections in the encoder-decoder structure.Compared with traditional short-time average/long-time average(STA/LTA)algorithms and advanced neural network models such as PhaseNet and EQTransformer,ResUNet shows superior performance in picking P-and S-wave phases.The model was trained on 400000 labeled microseismic signals from the Stanford earthquake dataset(STEAD)and was successfully applied to the Shizhuyuan polymetallic mine in Hunan Province,China.The results demonstrate that ResUNet achieves high picking accuracy and robustness in complex geological conditions,offering reliable technical support for early warning of disasters such as rockburst in deep underground engineering.
基金Project(SICGM2023301) supported by the State Key Laboratory of Strata Intelligent Control and Green Mining Co-founded by Shandong Province and the Ministry of Science and Technology,ChinaProject(SMDPC202202) supported by the Key Laboratory of Mining Disaster Prevention and Control,ChinaProject(U21A2030) supported by the National Natural Science Foundation of China。
文摘Microseismic (MS) source location plays an important role in MS monitoring. This paper proposes a MS source location method based on particle swarm optimization (PSO) and multi-sensor arrays, where a free weight joints the P-wave first arrival data. This method adaptively adjusts the preference for “superior” arrays and leverages “inferior” arrays to escape local optima, thereby improving the location accuracy. The effectiveness and stability of this method were validated through synthetic tests, pencil-lead break (PLB) experiments, and mining engineering applications. Specifically, for synthetic tests with 1 μs Gaussian noise and 100 μs large noise in rock samples, the location error of the multi-sensor arrays jointed location method is only 0.30 cm, which improves location accuracy by 97.51% compared to that using a single sensor array. The average location error of PLB events on three surfaces of a rock sample is reduced by 48.95%, 26.40%, and 55.84%, respectively. For mine blast event tests, the average location error of the dual sensor arrays jointed method is 62.74 m, 54.32% and 14.29% lower than that using only sensor arrays 1 and 2, respectively. In summary, the proposed multi-sensor arrays jointed location method demonstrates good noise resistance, stability, and accuracy, providing a compelling new solution for MS location in relevant mining scenarios.
基金This study is part of the project 2021RUAPON-REACT EU-Finanziamento PON“Ricerca e Innovazione”20142020,grant n.19-G-12543-2,funded by the Italian Ministry of University and Research(MUR)This study was developed in the frame of“The Geosciences for Sustainable Development”project(Budget Minis-tero dell'Universita e della Ricerca-Dipartimenti di Eccellenza 2023-2027,code n.C93C23002690001).
文摘A case study of seismic interferometry applied to a small microseismic monitoring network is here presented.The main objectives of this study are(i)to quantify the lateral variability of shear-wave ve-locities in the studied area,and(ii)to investigate the bias produced by noise directionality and non-stationarity in the velocity estimate.Despite the limited number of stations and the short-period char-acter of the seismic sensors,the empirical Green's functions were retrieved for all station pairs using two years of passive data.Both group and phase velocities were derived,the former using the widespread frequency-time analysis,the latter through the analysis of the real part of the cross-spectra.The main advantage of combining these two methods is a more accurate identification of higher modes,resulting in a reduction of ambiguity during picking and data interpretation.Surface wave tomography was run to obtain the spatial distribution of group and phase velocities for the same wavelengths.The low standard deviation of the results suggests that the sparse character of the network does not limit the applicability of the method,for this specific case.The obtained maps highlight the presence of a lower velocity area that extends from the centre of the network towards southeast.Group and phase velocity dispersion curves have been jointly inverted to retrieve as many shear-wave velocity profiles as selected station pairs.While the average model can be used for a more accurate location of the local natural seismicity,the associated standard deviations give us an indication of the lateral heterogeneity of seismic velocities as a function of depth.Finally,the same velocity analysis was repeated for different time windows in order to quantify the error associated to variations in the noise field.Errors as large as 4%have been found,related to the unfavorable orientation of the receiver pairs with respect to strongly directional noise sources,and to the very short time widows.It was shown that using a one-year time window these errors arereduced to 0.3%.
基金the financial support of the Key Technologies Research and Development Program(Grant No.2022YFC3003302)the National Natural Science Foundation of China(Grant Nos.51934007 and 52104230).
文摘Deep coal-energy mining frequently results in microseismic(MS)events,which may be a precursor to the risk of rockbursts and pose risks to human safety and infrastructure.Therefore,quantitatively predicting the time,energy,and location(TEL)of future MS events is crucial for understanding and preventing potential catastrophic events.In this study,we introduced the application of spatiotemporal graph convolutional networks(STGCN)to predict the TEL of MS events induced by deep coal-energy mining.Notably,this was the first application of graph convolution networks(GCNs)in the spatiotemporal prediction of MS events.The adjacency matrices of the sensor networks were determined based on the distance between MS sensors,the sensor network graphs we constructed,and GCN was employed to extract the spatiotemporal details of the graphs.The model is simple and versatile.By testing the model with on-site MS monitoring data,our results demonstrated promising efficacy in predicting the TEL of MS events,with the cosine similarity(C)above 0.90 and the mean relative error(MRE)below 0.08.This is critical to improving the safety and operational efficiency of deep coal-energy mining.
基金supported by the National Natural Science Foundation of China(Nos.42177173,U23A20651 and 42130719)and the Outstanding Youth Science Fund Project of Sichuan Provincial Natural Science Foundation(No.2025NSFJQ0003)。
文摘Dynamic stress adjustment in deep-buried high geostress hard rock tunnels frequently triggers catastrophic failures such as rockbursts and collapses.While a comprehensive understanding of this process is critical for evaluating surrounding rock stability,its dynamic evolution are often overlooked in engineering practice.This study systematically summarizes a novel classification framework for stress adjustment types—stabilizing(two-zoned),shallow failure(three-zoned),and deep failure(four-zoned)—characterized by distinct stress adjustment stages.A dynamic interpretation technology system is developed based on microseismic monitoring,integrating key microseismic parameters(energy index EI,apparent stressσa,microseismic activity S),seismic source parameter space clustering,and microseismic paths.This approach enables precise identification of evolutionary stages,stress adjustment types,and failure precursors,thereby elucidating the intrinsic linkage between geomechanical processes(stress redistribution)and failure risks.The study establishes criteria and procedures for identifying stress adjustment types and their associated failure risks,which were successfully applied in the Grand Canyon Tunnel of the E-han Highway to detect 50 instances of disaster risks.The findings offer invaluable insights into understanding the evolution process of stress adjustment and pinpointing the disaster risks linked to hard rock in comparable high geostress tunnels.
基金supported by National Natural Science Foundation of China Grant(No.42004040,42474092,U2239204,and 42304145)Natural Science Foundation of Jiangxi Province Grant(20242BAB25190 and 20232BAB213077).
文摘Accurate and rapid determination of source locations is of great significance for surface microseismic monitoring.Traditional methods,such as diffraction stacking,are time-consuming and challenging for real-time monitoring.In this study,we propose an approach to locate microseismic events using a deep learning algorithm with surface data.A fully convolutional network is designed to predict source locations.The input data is the waveform of a microseismic event,and the output consists of three 1D Gaussian distributions representing the probability distribution of the source location in the x,y,and z dimensions.The theoretical dataset is generated to train the model,and several data augmentation methods are applied to reduce discrepancies between the theoretical and field data.After applying the trained model to field data,the results demonstrate that our method is fast and achieves comparable location accuracy to the traditional diffraction stacking location method,making it promising for real-time microseismic monitoring.
基金supported by Earthquake Science and Technology Spark Program of China Earthquake Administration (No. XH20006Y)Local Standards Formulation and Revision Program of Hebei Province (No. FW202154)+1 种基金Earthquake Science and Technology Spark Program of Hebei Earthquake Agency (No. DZ2024112100002)2023 Seismological Data Sharing Project of China Earthquake Networks Center (Dataset Project)。
文摘In this study, we use the Bohai Sea area as an example to investigate the characteristics of secondary microseisms and their impact on seismic noise based on the temporal frequency spectral analysis of observation data from 33 broadband seismic stations during strong gust periods, and new perspectives are proposed on the generation mechanisms of secondary microseisms. The results show that short-period double- frequency (SPDF) and long-period double-frequency (LPDF) microseisms exhibit significant alternating trends of strengthening and weakening in the northwest area of the Bohai Sea. SPDF microseisms are generated by irregular wind waves during strong off shore wind periods, with a broad frequency band distributed in the range of 0.2-1 Hz;LPDF microseisms are generated by regular swells during periods of sea wind weakening, with a narrow frequency band concentrated between 0.15 and 0.3 Hz. In terms of temporal dimensions, as the sea wind weakens, the energy of SPDF microseisms weakens, and the dominant frequencies increase, whereas the energy of LPDF microseisms strengthens and the dominant frequencies decrease, which is consistent with the process of the decay of wind waves and the growth of swells. In terms of spatial dimensions, as the microseisms propagate inland areas, the advantageous frequency band and energy of SPDF microseisms are reduced and significantly attenuated, respectively, whereas LPDF microseisms show no significant changes. And during the propagation process in high-elevation areas, LPDF microseisms exhibit a certain site amplifi cation eff ect when the energy is strong. The results provide important supplements to the basic theory of secondary microseisms, preliminarily reveal the relationship between the atmosphere, ocean, and seismic noise, and provide important theoretical references for conducting geological and oceanographic research based on the characteristics of secondary microseisms.
基金financial support from the Distinguished Youth Funds of the National Natural Science Foundation of China(No.52425403)the Hunan Province Graduate Research Innovation Project of China(No.CX20230168)。
文摘The rock mass failure induced by deep mining exhibits pronounced spatial heterogeneity and diverse mechanisms,with its microseismic responses serving as effective indicators of regional failure evolution and instability mechanisms.Focusing on the Level VI stope sublayers in the Jinchuan#2 mining area,this study constructs a 24-parameter index system encompassing time-domain features,frequency-domain features,and multifractal characteristics.Through manifold learning,clustering analysis,and hybrid feature selection,15 key indicators were extracted to construct a classification framework for failure responses.Integrated with focal mechanism inversion and numerical simulation,the failure patterns and corresponding instability mechanisms across different structural zones were further identified.The results reveal that multiscale microseismic characteristics exhibit clear regional similarities.Based on the morphological features of radar plots derived from the 15 indicators,acoustic responses were classified into four typical types,each reflecting distinct local failure mechanisms,stress conditions,and plastic zone evolution.Moreover,considering dominant instability factors and rupture modes,four representative rock mass instability models were proposed for typical failure zones within the stope.These findings provide theoretical guidance and methodological support for hazard prediction,structural optimization,and disturbance control in deep metal mining areas.
基金supported by the National Natural Science Foundation of China(Grant Nos.U23A20297,52222810,52309126 and 52109116).
文摘In deeply buried tunneling projects,geological conditions are often complex and varied.Microseismic monitoring systems are extensively deployed to enhance construction safety.However,when the current geological conditions differ from those present during the signal collection for model training,recognition accuracy tends to decline significantly.Therefore,improving the applicability and stability of microseismic waveform recognition models across varying geological conditions has emerged as a critical challenge.To address this issue,we first analyze the impact of lithological changes and the development of structural planes on the features of microseismic waveforms.Subsequently,we propose a category-domain-aligned transfer learning method that enables the transfer of recognition capabilities across geological conditions by facilitating similar feature extraction and the recognition of cross-geological fracture waveforms.In this model,feature separation modeling enhances the extraction of category features of waveforms under different geological conditions.A deep transfer learning mechanism distinguishes between unique and common features,allowing for the capture of essential features necessary for model parameter updates.Through comparative experiments and feature distribution alignment and visualization,we demonstrate that the accuracy of microseismic waveform recognition across geological conditions achieves 90%.Additionally,the performance of our method is validated using microseismic signals collected from different sections of the construction site.
基金the financial support provided by the National Key Research and Development Program for Young Scientists(No.2021YFC2900400)Postdoctoral Fellowship Program of China Postdoctoral Science Foundation(CPSF)(No.GZB20230914)+2 种基金National Natural Science Foundation of China(No.52304123)China Postdoctoral Science Foundation(No.2023M730412)Chongqing Outstanding Youth Science Foundation Program(No.CSTB2023NSCQ-JQX0027).
文摘Monitoring sensors in complex engineering environments often record abnormal data,leading to significant positioning errors.To reduce the influence of abnormal arrival times,we introduce an innovative,outlier-robust localization method that integrates kernel density estimation(KDE)with damping linear correction to enhance the precision of microseismic/acoustic emission(MS/AE)source positioning.Our approach systematically addresses abnormal arrival times through a three-step process:initial location by 4-arrival combinations,elimination of outliers based on three-dimensional KDE,and refinement using a linear correction with an adaptive damping factor.We validate our method through lead-breaking experiments,demonstrating over a 23%improvement in positioning accuracy with a maximum error of 9.12 mm(relative error of 15.80%)—outperforming 4 existing methods.Simulations under various system errors,outlier scales,and ratios substantiate our method’s superior performance.Field blasting experiments also confirm the practical applicability,with an average positioning error of 11.71 m(relative error of 7.59%),compared to 23.56,66.09,16.95,and 28.52 m for other methods.This research is significant as it enhances the robustness of MS/AE source localization when confronted with data anomalies.It also provides a practical solution for real-world engineering and safety monitoring applications.
基金supported by the National Natural Science Foundation of China(Grant No.51934007)the Natural Science Foundation of Jiangsu Province,China(Grant No.BK20220691).
文摘Microseism,acoustic emission and electromagnetic radiation(M-A-E)data are usually used for predicting rockburst hazards.However,it is a great challenge to realize the prediction of M-A-E data.In this study,with the aid of a deep learning algorithm,a new method for the prediction of M-A-E data is proposed.In this method,an M-A-E data prediction model is built based on a variety of neural networks after analyzing numerous M-A-E data,and then the M-A-E data can be predicted.The predicted results are highly correlated with the real data collected in the field.Through field verification,the deep learning-based prediction method of M-A-E data provides quantitative prediction data for rockburst monitoring.
基金financial support of the Fundamental Research Funds for the Central Universities(Grant No.2022XSCX35)the National Natural Science Foundation of China(Grant Nos.51934007 and 52104230).
文摘Microseismic source location is crucial for the early warning of rockburst risks.However,the conventional methods face challenges in terms of the microseismic wave velocity and arrival time accuracy.Intelligent techniques,such as the full convolutional neural network(FCNN),can capture spatial information but struggle with complex microseismic sequence.Combining the FCNN with the long shortterm memory(LSTM)network enables better time-series signal classification by integrating multiscale information and is therefore suitable for waveform location.The LSTM-FCNN model does not require extensive data preprocessing and it simplifies the microseismic source location through feature extraction.In this study,we utilized the LSTM-FCNN as a regression learning model to locate the seismic focus.Initially,the method of short-time-average/long-time-average(STA/LTA)arrival time picking was employed to augment spatiotemporal information.Subsequently,oversampling the on-site data was performed to address the issue of data imbalance,and finally,the performance of LSTM-FCNN was tested.Meanwhile,we compared the LSTM-FCNN model with previous deep-learning models.Our results demonstrated remarkable location capabilities with a mean absolute error(MAE)of only 7.16 m.The model can realize swift training and high accuracy,thereby significantly improving risk warning of rockbursts.
基金supported by Western Research Interdisciplinary Initiative R6259A03.
文摘Rock fracture mechanisms can be inferred from moment tensors(MT)inverted from microseismic events.However,MT can only be inverted for events whose waveforms are acquired across a network of sensors.This is limiting for underground mines where the microseismic stations often lack azimuthal coverage.Thus,there is a need for a method to invert fracture mechanisms using waveforms acquired by a sparse microseismic network.Here,we present a novel,multi-scale framework to classify whether a rock crack contracts or dilates based on a single waveform.The framework consists of a deep learning model that is initially trained on 2400000+manually labelled field-scale seismic and microseismic waveforms acquired across 692 stations.Transfer learning is then applied to fine-tune the model on 300000+MT-labelled labscale acoustic emission waveforms from 39 individual experiments instrumented with different sensor layouts,loading,and rock types in training.The optimal model achieves over 86%F-score on unseen waveforms at both the lab-and field-scale.This model outperforms existing empirical methods in classification of rock fracture mechanisms monitored by a sparse microseismic network.This facilitates rapid assessment of,and early warning against,various rock engineering hazard such as induced earthquakes and rock bursts.
基金funded by the National Natural Science Foundation of China(Grant No.51874350)the National Natural Science Foundation of China(Grant No.52304127)+2 种基金the Fundamental Research Funds for the Central Universities of Central South University(Grant No.2020zzts200)the Science Foundation of the Fuzhou University(Grant No.511229)Fuzhou University Testing Fund of Precious Apparatus(Grant No.2024T040).
文摘The denoising of microseismic signals is a prerequisite for subsequent analysis and research.In this research,a new microseismic signal denoising algorithm called the Black Widow Optimization Algorithm(BWOA)optimized VariationalMode Decomposition(VMD)jointWavelet Threshold Denoising(WTD)algorithm(BVW)is proposed.The BVW algorithm integrates VMD and WTD,both of which are optimized by BWOA.Specifically,this algorithm utilizes VMD to decompose the microseismic signal to be denoised into several Band-Limited IntrinsicMode Functions(BLIMFs).Subsequently,these BLIMFs whose correlation coefficients with the microseismic signal to be denoised are higher than a threshold are selected as the effective mode functions,and the effective mode functions are denoised using WTD to filter out the residual low-and intermediate-frequency noise.Finally,the denoised microseismic signal is obtained through reconstruction.The ideal values of VMD parameters and WTD parameters are acquired by searching with BWOA to achieve the best VMD decomposition performance and solve the problem of relying on experience and requiring a large workload in the application of the WTD algorithm.The outcomes of simulated experiments indicate that this algorithm is capable of achieving good denoising performance under noise of different intensities,and the denoising performance is significantly better than the commonly used VMD and Empirical Mode Decomposition(EMD)algorithms.The BVW algorithm is more efficient in filtering noise,the waveform after denoising is smoother,the amplitude of the waveform is the closest to the original signal,and the signal-to-noise ratio(SNR)and the root mean square error after denoising are more satisfying.The case based on Fankou Lead-Zinc Mine shows that for microseismic signals with different intensities of noise monitored on-site,compared with VMD and EMD,the BVW algorithm ismore efficient in filtering noise,and the SNR after denoising is higher.
