Transformer-based models have significantly advanced binary code similarity detection(BCSD)by leveraging their semantic encoding capabilities for efficient function matching across diverse compilation settings.Althoug...Transformer-based models have significantly advanced binary code similarity detection(BCSD)by leveraging their semantic encoding capabilities for efficient function matching across diverse compilation settings.Although adversarial examples can strategically undermine the accuracy of BCSD models and protect critical code,existing techniques predominantly depend on inserting artificial instructions,which incur high computational costs and offer limited diversity of perturbations.To address these limitations,we propose AIMA,a novel gradient-guided assembly instruction relocation method.Our method decouples the detection model into tokenization,embedding,and encoding layers to enable efficient gradient computation.Since token IDs of instructions are discrete and nondifferentiable,we compute gradients in the continuous embedding space to evaluate the influence of each token.The most critical tokens are identified by calculating the L2 norm of their embedding gradients.We then establish a mapping between instructions and their corresponding tokens to aggregate token-level importance into instructionlevel significance.To maximize adversarial impact,a sliding window algorithm selects the most influential contiguous segments for relocation,ensuring optimal perturbation with minimal length.This approach efficiently locates critical code regions without expensive search operations.The selected segments are relocated outside their original function boundaries via a jump mechanism,which preserves runtime control flow and functionality while introducing“deletion”effects in the static instruction sequence.Extensive experiments show that AIMA reduces similarity scores by up to 35.8%in state-of-the-art BCSD models.When incorporated into training data,it also enhances model robustness,achieving a 5.9%improvement in AUROC.展开更多
Understanding dynamic visualization of mining-induced stress is of great significance to disaster prevention and control in coal mining activities.In this study,three theoretical models,including linear,polynomial,and...Understanding dynamic visualization of mining-induced stress is of great significance to disaster prevention and control in coal mining activities.In this study,three theoretical models,including linear,polynomial,and exponential models,are proposed to inverse the mining-induced stress through the acquisition and analysis of hydraulic support stress and micro-seismicity in the coal mining face.The distribution of mining-induced stress in the coal seam are graphed by fitting two key stress parameters including hydraulic support stress and peak stress,and two key zones including goaf zone and in situ stress zone.These key stress parameters and zones are defined based on the critical nodes of the model curve.According to the geological background of Mataihao coal mine in Erdos,Inner Mongolia Autonomous Region,China,the contours of mining-induced stress are graphed through the stress calculation of these three inversion theoretical models.The multi-monitoring data of micro-seismicity,drilling chips,advanced borehole stress and bolts axial force are used to verify the key stress parameters and zones of the theoretical models.It shows that the monitoring data are in good agreement with the distribution of inversed results.It should be emphasized that,if the fault structure exists around the mining face,the mining-induced stress decreases obviously when the mining face is passing through the faults,and the location of the peak stress will be closer to the mining face.The results in this study could provide methods for early prevention of extreme mining-induced stress and disaster control in the mining activities.展开更多
On December 18,2023,an M_(s)6.2 earthquake jolted Jishishan County in the Linxia Hui Autonomous Prefecture in Northwest China's Gansu Province,causing substantial casualties and building collapses.The earthquake o...On December 18,2023,an M_(s)6.2 earthquake jolted Jishishan County in the Linxia Hui Autonomous Prefecture in Northwest China's Gansu Province,causing substantial casualties and building collapses.The earthquake occurred in the Qilian Block on the northeastern border of the Qinghai-Tibet Plateau,where faults are highly active and the geological structure is complex.In this study,we utilized methods such as relocation,focal mechanism solutions,and earthquake rupture processes to describe seismogenic faults.The results indicated that the majority of aftershocks occurred at a depth of 12 km.The centroid depth of the main shock and the depth of the maximum rupture point during the rupture process were also 12 km.Various geophysical methods exhibited a high degree of consistency in depth exploration.Aftershocks were distributed mainly to the west and north of the main shock and extended in the NNW direction,primarily through unilateral rupture.The main shock was a reverse thrust event with a small dextral strike-slip component.In this study,more regional data,such as previous GPS observations,field geological observations,and the distributions of the primary stress states in the region,were also incorporated.We inferred that the main shock was triggered by the main fault at the northern margin of the Lajishan Fault and that the movement of the main fault also activated some secondary faults.The compressive forces on both sides of the Lajishan Fault Zone led to the uplift of mountain areas,accompanied by some landslides,leading to this catastrophic earthquake event.In this article,the activity relationships among the 2022 M_(s)6.9 Menyuan earthquake,the 2019 M_(s)5.7 Xiahe earthquake,and the Jishishan earthquake under the action of regional stress are also discussed.This study provides additional evidence and new ideas for exploring the seismogenic process of the Lajishan Fault Zone and has implications for future in-depth research on underground activity in this region.展开更多
The December 18,2023,M_(S)6.2 Jishishan earthquake occurred along the northeastern margin of the Qinghai-Xizang Plateau within the Laji-Jishi Shan Fault Zone(LJSFZ),a complex thrust-dominated tectonic belt.To identify...The December 18,2023,M_(S)6.2 Jishishan earthquake occurred along the northeastern margin of the Qinghai-Xizang Plateau within the Laji-Jishi Shan Fault Zone(LJSFZ),a complex thrust-dominated tectonic belt.To identify the seismogenic fault and better understand the regional tectonic framework,we integrated high-resolution Digital Elevation Models(DEMs)derived from GF-7 stereo satellite and Unmanned Aerial Vehicle(UAV)photogrammetry,relocated aftershock sequences,and conducted detailed field investigations.Our results identify four Late Quaternary thrust faults(F1-1 to F1-4),among which the Zhaomuchuan fault(F1-3),a NE-dipping back-thrust,aligns closely with the main-shock and aftershock distribution and exhibits clear Holocene activity.Seismic relocation reveals a NEdipping seismogenic zone at depths of 5-12 km,consistent with a shallow reverse-faulting mechanism under WSW-ENE oblique compression.Structural analysis and cross-sectional profiles suggest that fault F1-3 propagates into a mid-crustal detachment surface,forming a foreland-vergent thrust-nappe system.Importantly,the rupture of this secondary fault,rather than the locked primary boundary fault F1-1,indicates stress transfer and localization within a critically tapered wedge,consistent with global analogs of back-thrust-dominated earthquakes.These findings underscore the seismotectonic complexity of the LJSFZ and highlight the significant seismic hazard posed by subsidiary structures in compressional settings.展开更多
The widespread variation of focal depths and fault plane solutions observed in the Hindukush region depicts continuous deformation along the Indian-Eurasian collision zone.For period of twelve years i.e.from 2010 to 2...The widespread variation of focal depths and fault plane solutions observed in the Hindukush region depicts continuous deformation along the Indian-Eurasian collision zone.For period of twelve years i.e.from 2010 to 2022,a total of 89 intermediate-depth earthquakes of magnitude(Mw)≥5.5 of the Hindukush Region were considered,relocated using both regional and tele seismic data with 90 per cent confidence limits of less than 20 km.Two distinct seismic activity clusters:First one at a deeper depth and second at a shallower depth having different P-axes were observed that verifies the internal structure and geometry of Hindukush zone as suggested in previous studies.Beneath the Hindukush collision zone,there exists a complex pattern of deformation,arising from a combination of compression,tension,shearing and necking states due to an unusual and a rare case of subduction that is not from oceanic plate.The Hindukush seismic zone extends from 70 to 300 km depth and mostly strikes east-west and then turns northeast.The relocated seismicity by merging data of seismic network close to Hindukush along with international data shows that the Hindukush zone may be divided vertically into upper and lower slabs separated by a gap at about 150 km depth at which strike and dip directions change sharply with significant structural changes.Seismicity rate is higher in the lower part of Hindukush,having large magnitude events in a small volume below 180 km forming complex pattern of source mechanisms.Contrary in upper part seismicity rate is lower and scattered.The Global CMT(Global Centroid-Moment-Tensor Project)source mechanisms of intermediate depth earthquakes have a systematic pattern of reverse faulting with the vertical T-axes,while shallow events do not have such pattern.The vertical T-axes of the intermediate-depth events may be attributed to negative buoyancy caused by subduction of the cold and denser slab.展开更多
A new method, named relocation, was proposed to reduce the impact of sensor errors systematically, especially whenavailable data of sensors are abundant. The procedure includes evaluating the reliability of every sens...A new method, named relocation, was proposed to reduce the impact of sensor errors systematically, especially whenavailable data of sensors are abundant. The procedure includes evaluating the reliability of every sensors datum, processing the initiallocation by the credible data, and selecting a set of equations with optimal noise tolerance according to the relative relationshipbetween the initial location and sensors location, then calculating the final location by k-mean voting. The results obtained in thisresearch include comparing traditional location method with the presented method in both simulation and field experiment. In thefield experiment, the location error of relocation method reduced 41.8% compared with traditional location method. The resultssuggested that relocation method can improve the fault-tolerant performance significantly.展开更多
By investigating present relocation residential districts for peasants whose houses are removed for the unified planning of rural areas in north Jiangsu Province, as well as residents' feelings about the environme...By investigating present relocation residential districts for peasants whose houses are removed for the unified planning of rural areas in north Jiangsu Province, as well as residents' feelings about the environment of residential district, main architectural structures and energy consumption conditions, the indoor thermal environment, use of main heating and cooling facilities, residents' satisfaction on the acoustical and luminous environment, major space-enclosing structures and calculation of energy-saving designs are analyzed, and suggestions are given for the architectural design of relocation residential districts in the study area. It is stressed that the relationship between energy conservation and architectural layout, orientation, lighting, ventilation, selection of enclosing-structure materials, facade, color and style should be properly handled in the planning, and the focus is to control building orientation and shape coefficient, on the basis of which energy-saving designs of windows, exterior walls and roofs can be done. Energy consumption of present residential buildings is calculated and analyzed to bring forth new ideas to the energy-saving designs for relocation residential districts in north Jiangsu Province, and establish an architectural energy-saving system suitable for climatic and natural conditions of north Jiangsu to instruct the energy-saving designs of relocation residential districts in the study area.展开更多
An earthquake with Ms5.8 occurred on 10 March 2011 in Yingjiang county, western Yunnan, China. This earthquake caused 25 deaths and over 250 injuries. In order to better understand the seismotectonics in the region, w...An earthquake with Ms5.8 occurred on 10 March 2011 in Yingjiang county, western Yunnan, China. This earthquake caused 25 deaths and over 250 injuries. In order to better understand the seismotectonics in the region, we collected the arrival time data from the Yunnan seismic observational bulletins during 1 January to 25 March 2011, and precisely hand-picked the arrival times from high-quality seismograms that were recorded by the temporary seismic stations deployed by our Institute of Crustal Dynamics, China Earthquake Administration. Using these arrival times, we relocated all the earthquakes including the Yingjiang mainshock and its aftershocks using the double-difference relocation algorithm. Our results show that the relocated earthquakes dominantly occurred along the ENE direction and formed an upside-down bow-shaped structure in depth. It is also observed that after the Yingjiang mainshock, some aftershocks extended toward the SSE over about 10 km. These results may indicate that the Yingjiang mainshock ruptured a conjugate fault system consisting of the ENE trending Da Yingjiang fault and a SSE trending blind fault. Such structural features could contribute to severely seismic hazards during the moderate-size Yingjiang earthquake.展开更多
This paper analyses the impact of the poverty alleviation relocation(PAR)program on rural household income and evaluates the heterogeneous income effects of various relocation modes,based on a panel dataset of relocat...This paper analyses the impact of the poverty alleviation relocation(PAR)program on rural household income and evaluates the heterogeneous income effects of various relocation modes,based on a panel dataset of relocated households from 16 counties in eight Chinese provinces.The results show that participation in the PAR increases the income of both rural and urban resettlers.More specifically,it has a significant positive effect on agricultural and wage income for rural and urban resettlers,respectively.Further analyses show that the income increase for rural resettlers was mainly due to agricultural technology training and that the income increase for town resettlers was attributed to medical security.For the village resettlers,policies should focus on strengthening the development of local industries and training of agricultural technologies.For the urban resettlers,non-agricultural employment and public services in the urban resettlement areas should be promoted.展开更多
Micro-seismic phenomena, occurring when rock masses are subjected to forces and failures, allow the determination of their unstable states and failure zones by analyzing micro-seismic signals. We first present the pri...Micro-seismic phenomena, occurring when rock masses are subjected to forces and failures, allow the determination of their unstable states and failure zones by analyzing micro-seismic signals. We first present the principles of micro-seismic monitoring and location, as well as an underground explosion-proof micro-seismic monitoring system. Given a practical engineering application, we describe the application of micro-seismic monitoring technology in determining the height of a "two-zone" overburden, i.e., a caving zone and a fracture zone, the width of a coal-pillar section and the depth of failure of a floor. The workfaces monitored accomplished safe and highly efficient mining based on our micro-seismic monitoring results and provide direct proof of the reliability and validity of micro-seismic monitoring technology.展开更多
The medium-small earthquakes that occurred in the middle part of Tibetan Plateau(32°N–36°N, 90°E–93°E) from August 2016 to June 2017 were relocated using the absolute earthquake location method H...The medium-small earthquakes that occurred in the middle part of Tibetan Plateau(32°N–36°N, 90°E–93°E) from August 2016 to June 2017 were relocated using the absolute earthquake location method Hypo2000. Compared to the reports of Chinese Seismological Networks, our relocation results are more clustered on the whole, the horizontal location differences exceed 10 km, and the focal depths are concentrated in 0–8 km, which indicates that the upper crust inside the Tibetan Plateau is tectonically active. In June2017 altogether eight earthquakes above magnitude 3.0 took place; their relocated epicenters are concentrated around Gêladaindong.The relocation results of M<3.0 small earthquakes also showed obvious differences. Therefore, we used the CAP method to invert for the focal mechanisms of the M ≥3.0 earthquakes; results generally tally with the surface geological structures, indicating that the Tibetan Plateau is still under the strong compressional force from the India Plate. Among them the eight earthquakes that occurred near Gêladaindong in June 2017 are all of normal fault type or with some strike-slip at the same time; based on previous research results we conjecture that these events are intense shallow crust responses to deep crust-mantle activities.展开更多
The 2022 Menyuan M_(S)6.9 earthquake,which occurred on January 8,is the most destructive earthquake to occur near the Lenglongling(LLL)fault since the 2016 Menyuan M_(S)6.4 earthquake.We relocated the mainshock and af...The 2022 Menyuan M_(S)6.9 earthquake,which occurred on January 8,is the most destructive earthquake to occur near the Lenglongling(LLL)fault since the 2016 Menyuan M_(S)6.4 earthquake.We relocated the mainshock and aftershocks with phase arrival time observations for three days after the mainshock from the Qinghai Seismic Network using the double-difference method.The total length and width of the aftershock sequence are approximately 32 km and 5 km,respectively,and the aftershocks are mainly concentrated at a depth of 7-12 km.The relocated sequence can be divided into 18 km west and 13 km east segments with a boundary approximately 5 km east of the mainshock,where aftershocks are sparse.The east and west fault structures revealed by aftershock locations differ significantly.The west fault strikes EW and inclines to the south at a 71°-90°angle,whereas the east fault strikes 133°and has a smaller dip angle.Elastic strain accumulates at conjunctions of faults with different slip rates where it is prone to large earthquakes.Based on surface traces of faults,the distribution of relocated earthquake sequence and surface ruptures,the mainshock was determined to have occurred at the conjunction of the Tuolaishan(TLS)fault and LLL fault,and the west and east segments of the aftershock sequence were on the TLS fault and LLL fault,respectively.Aftershocks migrate in the early and late stages of the earthquake sequence.In the first 1.5 h after the mainshock,aftershocks expand westward from the mainshock.In the late stage,seismicity on the northeast side of the east fault is higher than that in other regions.The migration rate of the west segment of the aftershock sequence is approximately 4.5 km/decade and the afterslip may exist in the source region.展开更多
A relocation procedure to initialize tropical cyclones was developed to improve the representation of the initial conditions and the track forecast for Panasonic Weather Solutions Tropical Operational Forecasts. This ...A relocation procedure to initialize tropical cyclones was developed to improve the representation of the initial conditions and the track forecast for Panasonic Weather Solutions Tropical Operational Forecasts. This scheme separates the vortex perturbation and environment field from the first guess, then relocates the initial vortex perturbations to Lhe observed position by merging them with the environment field. The relationships of wind vector components with stream function and velocity potential are used for separating the vortex disturbance from first guess. For the separation of scalars, a low-pass Barnes filter is employed. The irregular-shaped relocation area corresponding to the specific initial conditions is determined by mapping the edge of the vortex radius in 36 directions.Then, the non-vortex perturbations in the relocation area are removed by a two-pass Barnes filter to retain the vortex perturbations, while the variable fields outside the perimeter of the modified vortex are kept ide.ntical to the original first guess. The potential impacts of this scheme on track forecasts were examined for three hurricane cases in the 2011-12 hurricane season. The experimental results demonstrate that the initialization scheme is able to effectively separate the vortex field from the environment field and maintain a relatively balanced and accurate relocated first guess. As the initial track error is reduced, the following track forecasts are considerably improved. The 72-h average track forecast error was redu,~ed by 32.6% for the cold-start cases, and by 38.4% when using the full-cycling data assimilation because of the accumulatedL improvements from the initialization scheme.展开更多
An Ms 6.4 earthquake occurred in Yangbi,Yunnan,China on May 21,2021,which has obvious foreshock activity and abundant aftershocks.Based on the seismic observation data recorded by the Yunnan Seismic Network three days...An Ms 6.4 earthquake occurred in Yangbi,Yunnan,China on May 21,2021,which has obvious foreshock activity and abundant aftershocks.Based on the seismic observation data recorded by the Yunnan Seismic Network three days before and seven days after the mainshock,a doubledifference location method was used to relocate 2133 earthquakes of the Yangbi sequence.Aftershocks are mostly distributed to the southeast of the mainshock in a unilateral rupture pattern.This sequence exhibits a SE-trending linear alignment with a length of about 25 km,and most of the focal depth is above 12 km.Integrated with the seismic distribution and focal mechanism results,we infer that the strike of the seismogenic fault is about 140°,and dipping to the SW.The fault structure revealed by the seismic sequence is complex,with the NW segment exhibiting a steep dip and relatively simple structure of strike-slip rupture and the SE segment consisting of several branching ruptures.The Yangbi Earthquake is a typical foreshock-mainshock-aftershock sequence,and the mainshock is likely triggered by the largest foreshock.This earthquake occurred in the boundary between high-and lowvelocity anomalous zone,where is susceptible to generate large earthquakes.展开更多
On January 10, 1998, an earthquake of ML=6.2 occurred in the border region between the Zhangbei County and Shangyi County of Hebei Province. This earthquake has been the most significant event occurred in the northern...On January 10, 1998, an earthquake of ML=6.2 occurred in the border region between the Zhangbei County and Shangyi County of Hebei Province. This earthquake has been the most significant event occurred in the northern China in the recent years. Historical seismicity in the Zhangbei-Shangyi region was very low. In the epicentral area no active fault capable of generating a moderate earthquake like this event was found. The earthquake locations of the main shock and its aftershocks of the Zhangbei-Shangyi earthquake sequence given by several agencies and authors were diverse and the resulted hypocentral distribution revealed no any dominant horizontal lineation. To study the seismogenic structure of the Zhangbei-Shangyi earthquake, in this paper the main shock and its aftershocks with ML3.0 of the Zhangbei-Shangyi earthquake sequence were relocated using the master event relative relocation algorithm. The relocated results show that the epicentral location of the main shock was 41.145癗, 114.462癊, which was located 4 km to the NE of the macro-epicenter of the main shock. The relocated focal depth of the main shock was 15 km. The hypocenters of the aftershocks distributed in a nearly vertical N20E-striking plane and its vicinity. The relocated results of the Zhangbei-Shangyi earthquake sequence clearly indicated that the seismogenic structure of this event was a nearly N-S- to NNE-SSW-striking fault with right-lateral and reverse slip, and that the occurrence of this event was associated with the horizontal and ENE-oriented compressive tectonic stress, which was compatible with the tectonic stress field in the northern China.展开更多
The failure depth of the coal seam floor is one important consideration that must be kept in mind when mining is carried out above a confined aquifer.Determining the floor failure depth is the essential precondition f...The failure depth of the coal seam floor is one important consideration that must be kept in mind when mining is carried out above a confined aquifer.Determining the floor failure depth is the essential precondition for predicting the water-resisting ability of the floor.We have used a high-precision microseismic monitoring technique to overcome the limited amount of data available from field measurements. The failure depth of a coal seam floor,especially an inclined coal seam floor,may be more accurately estimated by monitoring the continuous,dynamic failure of the floor.The monitoring results indicate the failure depth of the coal seam floor near the workface conveyance roadway（the lower crossheading） is deeper and that the failure range is wider here compared to the coal seam floor near the return airway（the upper crossheading）.The results of micro-seismic monitoring show that the dangerous area for water-inrush from the coal seam floor may be identified.This provides an important field measurement that helps ensure safe and highly efficient mining of the inclined coal seam above the confined aquifer at the Taoyuan Coal Mine.展开更多
A three-dimensional local-scale P-velocity model down to 25 km depth around the main shock epicenter region was constructed using 83821 event-to-receiver seismic rays from 5856 aftershocks recorded by a newly deployed...A three-dimensional local-scale P-velocity model down to 25 km depth around the main shock epicenter region was constructed using 83821 event-to-receiver seismic rays from 5856 aftershocks recorded by a newly deployed temporary seismic network. Checkerboard tests show that our tomographic model has lateral and vertical resolution of -2 km. The high-resolution P-velocity model revealed interesting structures in the seismogenic layer： （1） The Guanxian-Anxian fault, Yingxiu-Beichuan fault and Wenchuan-Maoxian fault of the Longmen Shan fault zone are well delineated by sharp upper crustal velocity changes; （2） The Pengguan massif has generally higher velocity than its surrounding areas, and may extend down to at least -10 km from the surface; （3） A sharp lateral velocity variation beneath the Wenchuan-Maoxian fault may indicate that the Pengguan massif＇s western boundary and/or the Wenchuan-Maoxian fault is vertical, and the hypocenter of the Wenchuan earthquake possibly located at the conjunction point of the NW dipping Yingxiu-Beichuan and Guanxian-Anxian faults, and vertical Wenchuan-Maoxian fault; （4） Vicinity along the Yingxiu- Beichuan fault is characterized by very low velocity and low seismicity at shallow depths, possibly due to high content of porosity and fractures; （5） Two blocks of low-velocity anomaly are respectively imaged in the hanging wall and foot wall of the Guanxian-Anxian fault with a -7 km offset with -5 km vertical component.展开更多
Blasting and breaking of hard roof are main inducing causes of rock bursts in coal mines with danger of rock burst,and it is important to find out the frequency spectrum distribution laws of these dynamic stress waves...Blasting and breaking of hard roof are main inducing causes of rock bursts in coal mines with danger of rock burst,and it is important to find out the frequency spectrum distribution laws of these dynamic stress waves and rock burst waves for researching the mechanism of rock burst.In this paper,Fourier transform as a micro-seismic signal conversion method of amplitude-time character to amplitude-frequency character is used to analyze the frequency spectrum characters of micro-seismic signal of blasting,hard roof breaking and rock bursts induced by the dynamic disturbance in order to find out the difference and relativity of different signals.The results indicate that blasting and breaking of hard roof are high frequency signals,and the peak values of dominant frequency of the signals are single.However,the results indicate that the rock bursts induced by the dynamic disturbance are low frequency signals,and there are two obvious peak values in the amplitude-frequency curve witch shows that the signals of rock bursts are superposition of low frequency signals and high frequency signals.The research conclusions prove that dynamic disturbance is necessary condition for rock bursts,and the conclusions provide a new way to research the mechanism of rock bursts.展开更多
An improved multidirectional velocity model was proposed for more accurately locating micro-seismic events in rock engineering. It was assumed that the stress wave propagation velocities from a micro-seismic source to...An improved multidirectional velocity model was proposed for more accurately locating micro-seismic events in rock engineering. It was assumed that the stress wave propagation velocities from a micro-seismic source to three nearest monitoring sensors in a sensor's array arrangement were the same. Since the defined objective function does not require pre-measurement of the stress wave propagation velocity in the field, errors from the velocity measurement can be avoided in comparison to three traditional velocity models. By analyzing 24 different cases, the proposed multidirectional velocity model iterated by the Simplex method is found to be the best option no matter the source is within the region of the sensor's array or not. The proposed model and the adopted iterative algorithm are verified by field data and it is concluded that it can significantly reduce the error of the estimated source location.展开更多
基金supported by Key Laboratory of Cyberspace Security,Ministry of Education,China。
文摘Transformer-based models have significantly advanced binary code similarity detection(BCSD)by leveraging their semantic encoding capabilities for efficient function matching across diverse compilation settings.Although adversarial examples can strategically undermine the accuracy of BCSD models and protect critical code,existing techniques predominantly depend on inserting artificial instructions,which incur high computational costs and offer limited diversity of perturbations.To address these limitations,we propose AIMA,a novel gradient-guided assembly instruction relocation method.Our method decouples the detection model into tokenization,embedding,and encoding layers to enable efficient gradient computation.Since token IDs of instructions are discrete and nondifferentiable,we compute gradients in the continuous embedding space to evaluate the influence of each token.The most critical tokens are identified by calculating the L2 norm of their embedding gradients.We then establish a mapping between instructions and their corresponding tokens to aggregate token-level importance into instructionlevel significance.To maximize adversarial impact,a sliding window algorithm selects the most influential contiguous segments for relocation,ensuring optimal perturbation with minimal length.This approach efficiently locates critical code regions without expensive search operations.The selected segments are relocated outside their original function boundaries via a jump mechanism,which preserves runtime control flow and functionality while introducing“deletion”effects in the static instruction sequence.Extensive experiments show that AIMA reduces similarity scores by up to 35.8%in state-of-the-art BCSD models.When incorporated into training data,it also enhances model robustness,achieving a 5.9%improvement in AUROC.
基金financially supported by the Independent Research fund of Joint National Local Engineering Research Centre for Safe and Precise Coal Mining(Anhui University of Science and Technology)(Grant No.EC2022001)State Key Research Development Program of China(Grant No.2022YFC3004602)the Fundamental Research Funds for the Central Universities(Grant No.2022YJSLJ08).
文摘Understanding dynamic visualization of mining-induced stress is of great significance to disaster prevention and control in coal mining activities.In this study,three theoretical models,including linear,polynomial,and exponential models,are proposed to inverse the mining-induced stress through the acquisition and analysis of hydraulic support stress and micro-seismicity in the coal mining face.The distribution of mining-induced stress in the coal seam are graphed by fitting two key stress parameters including hydraulic support stress and peak stress,and two key zones including goaf zone and in situ stress zone.These key stress parameters and zones are defined based on the critical nodes of the model curve.According to the geological background of Mataihao coal mine in Erdos,Inner Mongolia Autonomous Region,China,the contours of mining-induced stress are graphed through the stress calculation of these three inversion theoretical models.The multi-monitoring data of micro-seismicity,drilling chips,advanced borehole stress and bolts axial force are used to verify the key stress parameters and zones of the theoretical models.It shows that the monitoring data are in good agreement with the distribution of inversed results.It should be emphasized that,if the fault structure exists around the mining face,the mining-induced stress decreases obviously when the mining face is passing through the faults,and the location of the peak stress will be closer to the mining face.The results in this study could provide methods for early prevention of extreme mining-induced stress and disaster control in the mining activities.
基金funded by the National Natural Science Foundation of China(Grant No.42304072)。
文摘On December 18,2023,an M_(s)6.2 earthquake jolted Jishishan County in the Linxia Hui Autonomous Prefecture in Northwest China's Gansu Province,causing substantial casualties and building collapses.The earthquake occurred in the Qilian Block on the northeastern border of the Qinghai-Tibet Plateau,where faults are highly active and the geological structure is complex.In this study,we utilized methods such as relocation,focal mechanism solutions,and earthquake rupture processes to describe seismogenic faults.The results indicated that the majority of aftershocks occurred at a depth of 12 km.The centroid depth of the main shock and the depth of the maximum rupture point during the rupture process were also 12 km.Various geophysical methods exhibited a high degree of consistency in depth exploration.Aftershocks were distributed mainly to the west and north of the main shock and extended in the NNW direction,primarily through unilateral rupture.The main shock was a reverse thrust event with a small dextral strike-slip component.In this study,more regional data,such as previous GPS observations,field geological observations,and the distributions of the primary stress states in the region,were also incorporated.We inferred that the main shock was triggered by the main fault at the northern margin of the Lajishan Fault and that the movement of the main fault also activated some secondary faults.The compressive forces on both sides of the Lajishan Fault Zone led to the uplift of mountain areas,accompanied by some landslides,leading to this catastrophic earthquake event.In this article,the activity relationships among the 2022 M_(s)6.9 Menyuan earthquake,the 2019 M_(s)5.7 Xiahe earthquake,and the Jishishan earthquake under the action of regional stress are also discussed.This study provides additional evidence and new ideas for exploring the seismogenic process of the Lajishan Fault Zone and has implications for future in-depth research on underground activity in this region.
基金supported by the National Natural Science Foundation of China(Grant No.42277152,42041006)the Fundamental Research Funds for the Central Universities,Chang'an University(300102262910)+2 种基金supported by the International Science and Technology Cooperation Program of China(202406560140)Shaanxi Province Science and Technology Innovation Team(Ref.2021TD-51)the innovation team of Shaanxi Provincial Tri-Qin Scholars with Geoscience Big Data and Geohazard Prevention(2022)。
文摘The December 18,2023,M_(S)6.2 Jishishan earthquake occurred along the northeastern margin of the Qinghai-Xizang Plateau within the Laji-Jishi Shan Fault Zone(LJSFZ),a complex thrust-dominated tectonic belt.To identify the seismogenic fault and better understand the regional tectonic framework,we integrated high-resolution Digital Elevation Models(DEMs)derived from GF-7 stereo satellite and Unmanned Aerial Vehicle(UAV)photogrammetry,relocated aftershock sequences,and conducted detailed field investigations.Our results identify four Late Quaternary thrust faults(F1-1 to F1-4),among which the Zhaomuchuan fault(F1-3),a NE-dipping back-thrust,aligns closely with the main-shock and aftershock distribution and exhibits clear Holocene activity.Seismic relocation reveals a NEdipping seismogenic zone at depths of 5-12 km,consistent with a shallow reverse-faulting mechanism under WSW-ENE oblique compression.Structural analysis and cross-sectional profiles suggest that fault F1-3 propagates into a mid-crustal detachment surface,forming a foreland-vergent thrust-nappe system.Importantly,the rupture of this secondary fault,rather than the locked primary boundary fault F1-1,indicates stress transfer and localization within a critically tapered wedge,consistent with global analogs of back-thrust-dominated earthquakes.These findings underscore the seismotectonic complexity of the LJSFZ and highlight the significant seismic hazard posed by subsidiary structures in compressional settings.
文摘The widespread variation of focal depths and fault plane solutions observed in the Hindukush region depicts continuous deformation along the Indian-Eurasian collision zone.For period of twelve years i.e.from 2010 to 2022,a total of 89 intermediate-depth earthquakes of magnitude(Mw)≥5.5 of the Hindukush Region were considered,relocated using both regional and tele seismic data with 90 per cent confidence limits of less than 20 km.Two distinct seismic activity clusters:First one at a deeper depth and second at a shallower depth having different P-axes were observed that verifies the internal structure and geometry of Hindukush zone as suggested in previous studies.Beneath the Hindukush collision zone,there exists a complex pattern of deformation,arising from a combination of compression,tension,shearing and necking states due to an unusual and a rare case of subduction that is not from oceanic plate.The Hindukush seismic zone extends from 70 to 300 km depth and mostly strikes east-west and then turns northeast.The relocated seismicity by merging data of seismic network close to Hindukush along with international data shows that the Hindukush zone may be divided vertically into upper and lower slabs separated by a gap at about 150 km depth at which strike and dip directions change sharply with significant structural changes.Seismicity rate is higher in the lower part of Hindukush,having large magnitude events in a small volume below 180 km forming complex pattern of source mechanisms.Contrary in upper part seismicity rate is lower and scattered.The Global CMT(Global Centroid-Moment-Tensor Project)source mechanisms of intermediate depth earthquakes have a systematic pattern of reverse faulting with the vertical T-axes,while shallow events do not have such pattern.The vertical T-axes of the intermediate-depth events may be attributed to negative buoyancy caused by subduction of the cold and denser slab.
基金Projects(11472311,41272304,51504288)supported by the National Natural Science Foundation of China
文摘A new method, named relocation, was proposed to reduce the impact of sensor errors systematically, especially whenavailable data of sensors are abundant. The procedure includes evaluating the reliability of every sensors datum, processing the initiallocation by the credible data, and selecting a set of equations with optimal noise tolerance according to the relative relationshipbetween the initial location and sensors location, then calculating the final location by k-mean voting. The results obtained in thisresearch include comparing traditional location method with the presented method in both simulation and field experiment. In thefield experiment, the location error of relocation method reduced 41.8% compared with traditional location method. The resultssuggested that relocation method can improve the fault-tolerant performance significantly.
基金Key Project Process Mechanism and Prediction of Geological Hazards (2001CB711005-1-3) and State Key Basic Research Project Mechanism and Prediction of Continental Earthquakes (G1998040702). sponsored by the Ministry of Science and Techno
基金Supported by Talent-Introduction Scientific Research Program of Yancheng Institute of Technology(XKR2011078)~~
文摘By investigating present relocation residential districts for peasants whose houses are removed for the unified planning of rural areas in north Jiangsu Province, as well as residents' feelings about the environment of residential district, main architectural structures and energy consumption conditions, the indoor thermal environment, use of main heating and cooling facilities, residents' satisfaction on the acoustical and luminous environment, major space-enclosing structures and calculation of energy-saving designs are analyzed, and suggestions are given for the architectural design of relocation residential districts in the study area. It is stressed that the relationship between energy conservation and architectural layout, orientation, lighting, ventilation, selection of enclosing-structure materials, facade, color and style should be properly handled in the planning, and the focus is to control building orientation and shape coefficient, on the basis of which energy-saving designs of windows, exterior walls and roofs can be done. Energy consumption of present residential buildings is calculated and analyzed to bring forth new ideas to the energy-saving designs for relocation residential districts in north Jiangsu Province, and establish an architectural energy-saving system suitable for climatic and natural conditions of north Jiangsu to instruct the energy-saving designs of relocation residential districts in the study area.
基金supported by National Natural Science Foundation of China(Nos.40974201 and 40774044)to J.Lei
文摘An earthquake with Ms5.8 occurred on 10 March 2011 in Yingjiang county, western Yunnan, China. This earthquake caused 25 deaths and over 250 injuries. In order to better understand the seismotectonics in the region, we collected the arrival time data from the Yunnan seismic observational bulletins during 1 January to 25 March 2011, and precisely hand-picked the arrival times from high-quality seismograms that were recorded by the temporary seismic stations deployed by our Institute of Crustal Dynamics, China Earthquake Administration. Using these arrival times, we relocated all the earthquakes including the Yingjiang mainshock and its aftershocks using the double-difference relocation algorithm. Our results show that the relocated earthquakes dominantly occurred along the ENE direction and formed an upside-down bow-shaped structure in depth. It is also observed that after the Yingjiang mainshock, some aftershocks extended toward the SSE over about 10 km. These results may indicate that the Yingjiang mainshock ruptured a conjugate fault system consisting of the ENE trending Da Yingjiang fault and a SSE trending blind fault. Such structural features could contribute to severely seismic hazards during the moderate-size Yingjiang earthquake.
基金supported by the National Natural Science Foundation of China(71861147002 and 71761147004)supported by the Fundamental Research Funds for the Central Universitiesthe Research Funds of Renmin University of China(20XHN086)。
文摘This paper analyses the impact of the poverty alleviation relocation(PAR)program on rural household income and evaluates the heterogeneous income effects of various relocation modes,based on a panel dataset of relocated households from 16 counties in eight Chinese provinces.The results show that participation in the PAR increases the income of both rural and urban resettlers.More specifically,it has a significant positive effect on agricultural and wage income for rural and urban resettlers,respectively.Further analyses show that the income increase for rural resettlers was mainly due to agricultural technology training and that the income increase for town resettlers was attributed to medical security.For the village resettlers,policies should focus on strengthening the development of local industries and training of agricultural technologies.For the urban resettlers,non-agricultural employment and public services in the urban resettlement areas should be promoted.
基金supported by the National Basic Research Program of China (No. 2010CB226805)the National Natural Science Foundation of China (No. 50874103)+1 种基金the State Laboratory Fund (No. SKLGDUEK0905)the Talent Introduction Fund of Anhui University of Science & Technology
文摘Micro-seismic phenomena, occurring when rock masses are subjected to forces and failures, allow the determination of their unstable states and failure zones by analyzing micro-seismic signals. We first present the principles of micro-seismic monitoring and location, as well as an underground explosion-proof micro-seismic monitoring system. Given a practical engineering application, we describe the application of micro-seismic monitoring technology in determining the height of a "two-zone" overburden, i.e., a caving zone and a fracture zone, the width of a coal-pillar section and the depth of failure of a floor. The workfaces monitored accomplished safe and highly efficient mining based on our micro-seismic monitoring results and provide direct proof of the reliability and validity of micro-seismic monitoring technology.
基金supported by the National Key R/D Project (2016YFC0600301, 2016YFE0109300)the National Natural Science Foundation of China (41574086, 41761134094)China Geological Survey project (DD20160022-05)
文摘The medium-small earthquakes that occurred in the middle part of Tibetan Plateau(32°N–36°N, 90°E–93°E) from August 2016 to June 2017 were relocated using the absolute earthquake location method Hypo2000. Compared to the reports of Chinese Seismological Networks, our relocation results are more clustered on the whole, the horizontal location differences exceed 10 km, and the focal depths are concentrated in 0–8 km, which indicates that the upper crust inside the Tibetan Plateau is tectonically active. In June2017 altogether eight earthquakes above magnitude 3.0 took place; their relocated epicenters are concentrated around Gêladaindong.The relocation results of M<3.0 small earthquakes also showed obvious differences. Therefore, we used the CAP method to invert for the focal mechanisms of the M ≥3.0 earthquakes; results generally tally with the surface geological structures, indicating that the Tibetan Plateau is still under the strong compressional force from the India Plate. Among them the eight earthquakes that occurred near Gêladaindong in June 2017 are all of normal fault type or with some strike-slip at the same time; based on previous research results we conjecture that these events are intense shallow crust responses to deep crust-mantle activities.
基金jointly funded by the National Key Research and Development Program of China (No. 2021YFC3000702)the Special Fund of the Institute of Geophysics, China Earthquake Administration (No. DQJB21Z05)the National Natural Science Foundation of China (No. 41804062)
文摘The 2022 Menyuan M_(S)6.9 earthquake,which occurred on January 8,is the most destructive earthquake to occur near the Lenglongling(LLL)fault since the 2016 Menyuan M_(S)6.4 earthquake.We relocated the mainshock and aftershocks with phase arrival time observations for three days after the mainshock from the Qinghai Seismic Network using the double-difference method.The total length and width of the aftershock sequence are approximately 32 km and 5 km,respectively,and the aftershocks are mainly concentrated at a depth of 7-12 km.The relocated sequence can be divided into 18 km west and 13 km east segments with a boundary approximately 5 km east of the mainshock,where aftershocks are sparse.The east and west fault structures revealed by aftershock locations differ significantly.The west fault strikes EW and inclines to the south at a 71°-90°angle,whereas the east fault strikes 133°and has a smaller dip angle.Elastic strain accumulates at conjunctions of faults with different slip rates where it is prone to large earthquakes.Based on surface traces of faults,the distribution of relocated earthquake sequence and surface ruptures,the mainshock was determined to have occurred at the conjunction of the Tuolaishan(TLS)fault and LLL fault,and the west and east segments of the aftershock sequence were on the TLS fault and LLL fault,respectively.Aftershocks migrate in the early and late stages of the earthquake sequence.In the first 1.5 h after the mainshock,aftershocks expand westward from the mainshock.In the late stage,seismicity on the northeast side of the east fault is higher than that in other regions.The migration rate of the west segment of the aftershock sequence is approximately 4.5 km/decade and the afterslip may exist in the source region.
基金jointly supported by National Center for Atmospheric Research(NCAR)and Panasonic Avionics CorporationNational Natural Science Foundation of China(NSFC)under Grant No.40975068+1 种基金the National Basic Research Program of China(973 Program)under Grand No.2013CB430102Jiangsu province colleges and universities natural science major basic research projects
文摘A relocation procedure to initialize tropical cyclones was developed to improve the representation of the initial conditions and the track forecast for Panasonic Weather Solutions Tropical Operational Forecasts. This scheme separates the vortex perturbation and environment field from the first guess, then relocates the initial vortex perturbations to Lhe observed position by merging them with the environment field. The relationships of wind vector components with stream function and velocity potential are used for separating the vortex disturbance from first guess. For the separation of scalars, a low-pass Barnes filter is employed. The irregular-shaped relocation area corresponding to the specific initial conditions is determined by mapping the edge of the vortex radius in 36 directions.Then, the non-vortex perturbations in the relocation area are removed by a two-pass Barnes filter to retain the vortex perturbations, while the variable fields outside the perimeter of the modified vortex are kept ide.ntical to the original first guess. The potential impacts of this scheme on track forecasts were examined for three hurricane cases in the 2011-12 hurricane season. The experimental results demonstrate that the initialization scheme is able to effectively separate the vortex field from the environment field and maintain a relatively balanced and accurate relocated first guess. As the initial track error is reduced, the following track forecasts are considerably improved. The 72-h average track forecast error was redu,~ed by 32.6% for the cold-start cases, and by 38.4% when using the full-cycling data assimilation because of the accumulatedL improvements from the initialization scheme.
基金financially supported by National Science Foundation of China(No.41774067)the National Key R&D Program of China(No.2018YFC1503400)+1 种基金the Special Fund of the Institute of GeophysicsChina Earthquake Administration(No.DQJB20X07)。
文摘An Ms 6.4 earthquake occurred in Yangbi,Yunnan,China on May 21,2021,which has obvious foreshock activity and abundant aftershocks.Based on the seismic observation data recorded by the Yunnan Seismic Network three days before and seven days after the mainshock,a doubledifference location method was used to relocate 2133 earthquakes of the Yangbi sequence.Aftershocks are mostly distributed to the southeast of the mainshock in a unilateral rupture pattern.This sequence exhibits a SE-trending linear alignment with a length of about 25 km,and most of the focal depth is above 12 km.Integrated with the seismic distribution and focal mechanism results,we infer that the strike of the seismogenic fault is about 140°,and dipping to the SW.The fault structure revealed by the seismic sequence is complex,with the NW segment exhibiting a steep dip and relatively simple structure of strike-slip rupture and the SE segment consisting of several branching ruptures.The Yangbi Earthquake is a typical foreshock-mainshock-aftershock sequence,and the mainshock is likely triggered by the largest foreshock.This earthquake occurred in the boundary between high-and lowvelocity anomalous zone,where is susceptible to generate large earthquakes.
基金The Project Mechanism and Prediction of Continental Strong Earthquakes Ministry of Science and Technology Peoples Republic of China (G19980407/95-13-02-04).
文摘On January 10, 1998, an earthquake of ML=6.2 occurred in the border region between the Zhangbei County and Shangyi County of Hebei Province. This earthquake has been the most significant event occurred in the northern China in the recent years. Historical seismicity in the Zhangbei-Shangyi region was very low. In the epicentral area no active fault capable of generating a moderate earthquake like this event was found. The earthquake locations of the main shock and its aftershocks of the Zhangbei-Shangyi earthquake sequence given by several agencies and authors were diverse and the resulted hypocentral distribution revealed no any dominant horizontal lineation. To study the seismogenic structure of the Zhangbei-Shangyi earthquake, in this paper the main shock and its aftershocks with ML3.0 of the Zhangbei-Shangyi earthquake sequence were relocated using the master event relative relocation algorithm. The relocated results show that the epicentral location of the main shock was 41.145癗, 114.462癊, which was located 4 km to the NE of the macro-epicenter of the main shock. The relocated focal depth of the main shock was 15 km. The hypocenters of the aftershocks distributed in a nearly vertical N20E-striking plane and its vicinity. The relocated results of the Zhangbei-Shangyi earthquake sequence clearly indicated that the seismogenic structure of this event was a nearly N-S- to NNE-SSW-striking fault with right-lateral and reverse slip, and that the occurrence of this event was associated with the horizontal and ENE-oriented compressive tectonic stress, which was compatible with the tectonic stress field in the northern China.
基金supported by the National Basic Research Program ofChina(No.2010CB202210)the National Natural Science Foundation of China(No.50874103)+1 种基金the Natural Science Foundation of Jiangsu Province(No.KB2008135)as well as by the Qinglan Project of Jiangsu Province
文摘The failure depth of the coal seam floor is one important consideration that must be kept in mind when mining is carried out above a confined aquifer.Determining the floor failure depth is the essential precondition for predicting the water-resisting ability of the floor.We have used a high-precision microseismic monitoring technique to overcome the limited amount of data available from field measurements. The failure depth of a coal seam floor,especially an inclined coal seam floor,may be more accurately estimated by monitoring the continuous,dynamic failure of the floor.The monitoring results indicate the failure depth of the coal seam floor near the workface conveyance roadway（the lower crossheading） is deeper and that the failure range is wider here compared to the coal seam floor near the return airway（the upper crossheading）.The results of micro-seismic monitoring show that the dangerous area for water-inrush from the coal seam floor may be identified.This provides an important field measurement that helps ensure safe and highly efficient mining of the inclined coal seam above the confined aquifer at the Taoyuan Coal Mine.
基金the Basic Research Foundation of the Institute of Geomechanics,CAGS(grant DZLXJK200707)Natural Science Foundation of China grant 40674058)
文摘A three-dimensional local-scale P-velocity model down to 25 km depth around the main shock epicenter region was constructed using 83821 event-to-receiver seismic rays from 5856 aftershocks recorded by a newly deployed temporary seismic network. Checkerboard tests show that our tomographic model has lateral and vertical resolution of -2 km. The high-resolution P-velocity model revealed interesting structures in the seismogenic layer： （1） The Guanxian-Anxian fault, Yingxiu-Beichuan fault and Wenchuan-Maoxian fault of the Longmen Shan fault zone are well delineated by sharp upper crustal velocity changes; （2） The Pengguan massif has generally higher velocity than its surrounding areas, and may extend down to at least -10 km from the surface; （3） A sharp lateral velocity variation beneath the Wenchuan-Maoxian fault may indicate that the Pengguan massif＇s western boundary and/or the Wenchuan-Maoxian fault is vertical, and the hypocenter of the Wenchuan earthquake possibly located at the conjunction point of the NW dipping Yingxiu-Beichuan and Guanxian-Anxian faults, and vertical Wenchuan-Maoxian fault; （4） Vicinity along the Yingxiu- Beichuan fault is characterized by very low velocity and low seismicity at shallow depths, possibly due to high content of porosity and fractures; （5） Two blocks of low-velocity anomaly are respectively imaged in the hanging wall and foot wall of the Guanxian-Anxian fault with a -7 km offset with -5 km vertical component.
基金the National Basic Research Program of China (Nos.2005 CB221504 and 2010CB226805)the Research Fund of the State Key Laboratory of Coal Resources and Mine Safety,CUMT (No.09KF08)the Foundation of the Henan Educational Committee (No.2010 A440003)
文摘Blasting and breaking of hard roof are main inducing causes of rock bursts in coal mines with danger of rock burst,and it is important to find out the frequency spectrum distribution laws of these dynamic stress waves and rock burst waves for researching the mechanism of rock burst.In this paper,Fourier transform as a micro-seismic signal conversion method of amplitude-time character to amplitude-frequency character is used to analyze the frequency spectrum characters of micro-seismic signal of blasting,hard roof breaking and rock bursts induced by the dynamic disturbance in order to find out the difference and relativity of different signals.The results indicate that blasting and breaking of hard roof are high frequency signals,and the peak values of dominant frequency of the signals are single.However,the results indicate that the rock bursts induced by the dynamic disturbance are low frequency signals,and there are two obvious peak values in the amplitude-frequency curve witch shows that the signals of rock bursts are superposition of low frequency signals and high frequency signals.The research conclusions prove that dynamic disturbance is necessary condition for rock bursts,and the conclusions provide a new way to research the mechanism of rock bursts.
基金Project(IRT0950)supported by the Cheung Kong Scholars and the Development Plan of Innovative Team,ChinaProject supported by China Scholarship Council
文摘An improved multidirectional velocity model was proposed for more accurately locating micro-seismic events in rock engineering. It was assumed that the stress wave propagation velocities from a micro-seismic source to three nearest monitoring sensors in a sensor's array arrangement were the same. Since the defined objective function does not require pre-measurement of the stress wave propagation velocity in the field, errors from the velocity measurement can be avoided in comparison to three traditional velocity models. By analyzing 24 different cases, the proposed multidirectional velocity model iterated by the Simplex method is found to be the best option no matter the source is within the region of the sensor's array or not. The proposed model and the adopted iterative algorithm are verified by field data and it is concluded that it can significantly reduce the error of the estimated source location.
