1643 results about "Seismic wave" patented technology

The invention belongs to the technical field of earthquake disaster scene recognition, and discloses an earthquake disaster scene recognition method based on deep learning. The method comprises the following specific steps: S1, data acquisition and preprocessing; S1.1, multi-source heterogeneous data acquisition and establishment of a comprehensive database containing seismic waveform data, surface deformation data, building structure data, geographic information data and historical disaster record data; through fusion of a 3D convolutional network, a graph attention mechanism, a space-time LSTM and an adaptive cross-modal attention fusion technology, combined modeling of a seismic waveform space-time evolution law, an earth surface deformation space distribution characteristic, a building group topology vulnerability and disaster chain time sequence association is realized, the characterization capability of a complex nonlinear disaster mode is effectively improved, and the method has the advantages of high adaptability and high reliability. And disaster assessment response time is shortened to a sub-second level through mixed precision quantification and edge computing deployment, and high recognition accuracy is still kept in a scene with strong noise and data missing in combination with a multi-task classifier and a physical constraint verification mechanism.

The invention discloses an Internet of Things multichannel earthquake early warning emergency linkage system, belongs to the technical field of earthquake early warning, and aims to solve the problems of insufficient real-time performance, reliability, scene coverage and safety in the prior art. The sensing layer collects data cooperatively through various sensors, eliminates environmental interference and ensures high-precision seismic wave detection, the edge calculation layer filters noise in real time by adopting an advanced algorithm, and confirms seismic events through multi-dimensional comparison, so that the local processing efficiency is remarkably improved, the cloud load is greatly reduced, a cloud platform fuses multi-source data, and the seismic detection efficiency is improved. A machine learning model is combined to dynamically optimize an early warning threshold value, the accuracy of earthquake type classification and intensity prediction is improved, full-link optimization from data acquisition to intelligent decision making is realized, a multi-channel distribution module is designed, instructions are transmitted according to priority levels, and urban and rural full-scene coverage is ensured through multiple communication technologies.

The invention discloses a tunnel deep-buried drainage ditch rapid blasting excavation method based on vertical drilling, which comprises the following steps: step 1, detecting the geology in front of a tunnel by using vertical drilling coring, geological radar and TSP seismic wave technology, and constructing a three-dimensional model; 2, generating an asymmetric hole net layout based on a geological model and AI, and setting a drilling-free vibration isolation belt; the design of dense inside and sparse outside is adopted, and vibration propagation is blocked; thirdly, a three-stage air spacer is arranged at the bottom of the blast hole, segmented charging is implemented, the explosive energy release time is prolonged, and the single-segment explosive quantity peak value is reduced; fourthly, an electronic detonator is adopted for millisecond detonation, and peak vibration is weakened through the waveform superposition offset effect; 5, arranging fiber grating sensors in the tunnel, collecting data in real time and transmitting the data to the cloud AI platform; dynamically adjusting parameters based on the monitoring data; and 6, evaluating a blasting effect through a three-dimensional laser scanner, generating an evaluation report in combination with data, updating a BIM model database, and continuously optimizing an AI algorithm.

The invention provides a tunnel construction structure health on-line monitoring method and system, and the method comprises the steps: obtaining stress data, crack images and seismic wave signals of tunnel construction, and carrying out the preprocessing; obtaining and constructing stress-crack joint representation mapping, generating a first joint feature vector, and predicting a lining health state of a future # imgabs0 # time step; generating a second joint feature vector, and determining a stratum stress field of the future # imgabs1 # time step according to the second joint feature vector; constructing a tunnel lining damage probability model according to the stress data and the crack image, and determining an overall construction risk score; and dividing the construction risks into a plurality of grades. According to the method, data multi-modal fusion, structure-stratum joint modeling and AI enhanced prediction and early warning are achieved, the defects in the prior art are effectively overcome, the method can be widely applied to engineering scenes such as tunnel construction, subway construction and underground space development, and the construction safety and the structure health monitoring capacity are remarkably improved.

The invention discloses a seismic data constrained direct current method three-dimensional inversion method, which relates to the field of geophysical exploration and comprises the steps of processing seismic data and performing inversion processing on resistivity data. The seismic wave layer velocity information is used for constraining high-density electrical method inversion, so that a supplement effect to a certain degree can be achieved, three-dimensional seismic image information can be used for supplementing the resolution ratio of resistivity inversion, the boundary of an underground electrical anomalous body is reflected, and meanwhile, through the comprehensive geological-geophysical information, the electrical anomalous body can be accurately determined. And boundary features of the anomalous body can be further depicted. Seismic data constraints are applied to three-dimensional resistivity inversion, numerical simulation and inversion processing are performed on electrical method data under rugged topography according to finite volume method numerical simulation and a Gaussian-Newton inversion method, and meanwhile, seismic anisotropic velocity tensor constraints are introduced to perform inversion, so that the problem of multiplicity of solutions of inversion of a single method is reduced, and the inversion accuracy is improved. A more accurate three-dimensional underground electrical structure is obtained, and the resolution of inversion to a longitudinal boundary is improved.

The invention discloses a road protection multi-scene coupling simulation method and system based on digital twinning, and relates to the field of electric digital data processing, and the method comprises the steps: collecting dynamic response data (strain, acceleration and displacement) of a traffic protection facility through multiple sensors, and building a scene feature library containing vehicle collision, wind load and seismic waves; and projecting response data to a unified time axis, identifying a multi-scene overlapping interval, and determining an action sequence and duration. Primary and secondary scenes are judged according to scene action parameters, and mechanical correction values and standard response parameters are extracted. And establishing a digital twinborn model for coupling simulation to obtain a stress distribution condition, evaluating a structural damage level and generating a safety performance report. By implementing the method, the deviation between the mechanical property simulation result of the protection facility and the actual working condition can be reduced.

The invention relates to the technical field of building information, and discloses an ancient building anti-seismic performance analysis and evaluation system based on BIM modeling, and the system comprises a building parameter obtaining module which is used for dynamically collecting component material attributes, structure connection states and historical damage parameters, recognizing structure weak links, and triggering a collaborative evaluation instruction; the modal feature analysis module and the seismic response simulation module are used for respectively extracting vibration modal feature parameters and seismic wave energy distribution parameters to obtain modal feature integrity indexes and dynamic stability evaluation values; the multi-source fusion judgment module synthesizes the indexes to generate a structure reinforcing or component repairing signal; the optimization parameter generation module is matched with the anti-seismic optimization strategy and generates structure reinforcement and node connection repair parameters; the system further comprises a damage coupling analysis module and a model iteration updating module, and monitoring of historical damage evolution and closed-loop updating of the BIM model are achieved. According to the system, precise evaluation and dynamic optimization of the anti-seismic performance of the historic building are realized.

The invention provides a single-station earthquake positioning system and method combining multi-source features and collaborative loss, and relates to the technical field of earthquake positioning, and the single-station earthquake positioning system comprises a data preprocessing module, a model training module, and a single seismic source positioning model. Inputting three-component waveform data into a data preprocessing module by querying seismic directory information, inputting the preprocessed seismic waveform data into a single seismic source positioning model, training each network module to obtain a single-station seismic positioning system, and outputting a time sequence feature tensor, a time frequency feature tensor and a station information feature tensor; the trained seismic positioning system is used for predicting the seismic source depth and the epicentral distance, seismic waveform data collected by the station are input into the data preprocessing module to obtain standardized waveform data, the standardized waveform data are input into the seismic source positioning module, and prediction results of the seismic source depth and the epicentral distance of the station are output.

The invention discloses a data center seismic isolation performance detection method based on AI intelligence, and particularly relates to the technical field of seismic isolation performance detection. Earthquake vibration characteristic data are collected in real time on a data center building and equipment, earthquake input characteristics are extracted in combination with a historical earthquake wave database, a seismic isolation performance prediction model is constructed and trained, a foundation shear wave velocity gradient abnormal value and a seismic isolation layer dynamic displacement response deviation ratio are further extracted, a deviation index is comprehensively calculated, and a seismic isolation performance prediction model is constructed and trained. And classification processing is carried out according to the deviation level, especially under the condition of a middle deviation level, foundation physical parameters are dynamically updated, and the model is optimized, so that the accuracy of the prediction model and the self-adaptive capability of the system are remarkably improved, and the shock isolation safety and the continuous operation capability of the data center in a strong shock environment are effectively guaranteed.

The invention provides a rainfall and earthquake coupled slope simulation method and system. The method comprises the following steps: constructing a coupling numerical model through multi-physics field data of a slope; the dynamic coupling effect of rainfall and earthquake is simulated by initializing a seepage field and loading earthquake power; the stability of the simulation result is comprehensively evaluated through iteration of a strength reduction method, and whole-process coupling from seepage to dynamic response of the slope is achieved. SeismoSignal is used for processing seismic waves, and the problem that the value of kinematics parameters is not zero when the seismic action is finished is avoided; the allowable maximum negative pore water pressure boundary is assigned to the slope surface of the side slope, the actual steady-state seepage field of the side slope is restored, and the seepage field in the side slope under different rainfall conditions is accurately calculated; by adopting an unsaturated soil dynamic seepage model for a seepage field, Rayleigh damping is introduced and an earthquake load is applied, so that cross-time scale coupling is realized. The method is simple and convenient to operate, high in calculation efficiency and accurate in calculation result, and has practical significance and value in research of the problem in the rock-soil field.

The invention relates to the technical field of petroleum and natural gas engineering, and discloses an oil and gas reservoir optimized mining method based on multi-modal data, which comprises the following steps: constructing an oil and gas reservoir multi-modal data acquisition system, seismic wave field data, logging interpretation data, production dynamic data, micro-seismic monitoring data, underground temperature and pressure time sequence data and shaft structure parameter data are obtained through the acquisition system; and performing space-time alignment processing on the acquired multi-modal data, establishing a unified geological coordinate system and a time reference, and eliminating data isomerism caused by different acquisition frequencies and spatial resolutions. A six-dimensional heterogeneous data acquisition system covering a seismic wave field, well logging interpretation, production dynamics, micro-seismic monitoring, an underground temperature and pressure time sequence and shaft structure parameters is constructed, so that the depiction precision of a reservoir porosity field, a permeability field, a saturation field and a pressure field is essentially improved.

A method for monitoring a rock burst in ultra-thin coal seams, which method belongs to the technical field of fully mechanized coal mining in ultra-thin coal seams. The method comprises: respectively recording the acquired propagation times and propagation velocities of attenuated seismic waves from artificial seismic sources and the acquired propagation time and propagation velocity of a rock burst, and comparing different seismic wave signals and real-time rock burst signals passing through a mining area; converting the different seismic wave signals and the real-time rock burst signals into digital signals, and determining via comparison whether the digital signals have waveforms during the rock burst; when the digital signals have waveforms during the rock burst, identifying the waveforms of micro-seismic events, and determining the energy intensity of the rock burst; locating rock burst sources, and obtaining an intersection area of the rock burst sources, so as to obtain a seismic source position after a delineated range is reduced; and on the basis of acquired real-time pressure data in underground rock or coal seams, obtaining the positions of the maximum pressure and maximum pressure changes of a seismic source, and the numerical values thereof by means of analyzing a real-time pressure value and a real-time strain change rate. By means of the method, potential hazardous areas can be monitored. The present invention further relates to a system for monitoring a rock burst in ultra-thin coal seams.

The invention relates to the technical field of geophysical exploration, in particular to an exploration surveying and mapping method based on remote sensing and a surveying and mapping instrument thereof. The exploration surveying and mapping method based on remote sensing comprises the following steps: step 1, multi-platform collaborative data dynamic acquisition; step 2, multi-modal data fusion and feature enhancement: vegetation noise filtering is performed on the laser radar point cloud data, and a surface terrain model under vegetation coverage is reconstructed; seismic wave speed data and ground penetrating radar dielectric constant data are combined to construct an underground medium physical property combined distribution model; associating the reflectivity characteristics of the multispectral image with the underground physical property parameters, and identifying a geological abnormal region; and step 3, hidden geologic body dynamic modeling and risk assessment. By means of the method, real-time monitoring and risk assessment of the complex geologic body can be achieved, and the geological disaster prediction and prevention and control capacity is greatly improved.

The invention discloses a semi-supervised seismic wave impedance inversion method and system based on Swin Transform iterative gradient correction, and belongs to the technical field of earthquakes in geophysical exploration. According to the method, the impedance gradient is iteratively corrected in the model space through the deep neural network, and complex nonlinear prior information is captured; and each anisotropic structure direction regularization item is established, and constraint is carried out along the direction of the stratum structure, so that the space consistency of the prediction result is ensured. Meanwhile, the USTNet network designed by the invention combines the advantages of the local feature capture capability of the convolutional neural network and the global modeling capability of Transform, and reduces the non-uniqueness of solutions. The network is used for inverting synthetic data and actual data, and the effectiveness of the method is verified.

The invention discloses a deep learning seismic source mechanism inversion system and method based on physical driving. The system is characterized in that a seismic wave acquisition module is used for acquiring an actually measured seismic wave waveform; the data preprocessing module preprocesses the data; the data feature extractor is used for extracting seismic waveform data features; the physical constraint optimizer is used for constructing a comprehensive loss item function; the model training module is used for training the seismic source mechanism prediction model; the seismic source mechanism prediction module is used for result inversion; the method comprises the following steps: simulating and generating a synthetic seismic waveform, and extracting spectrum features; preprocessing the data; extracting multi-level enhanced features from actually measured and synthesized seismic waveforms by using a data feature extractor; introducing the fault physical mechanism, the stress tensor conservation and the seismic source tensor constraint into a loss function to form a comprehensive loss item, and constructing a seismic source mechanism prediction model; and performing inversion of an actual seismic event by using the seismic source mechanism prediction model. According to the system and the method, efficient and accurate seismic source mechanism inversion can be carried out on unknown seismic events.

The invention discloses an urban karst water guide channel detection system and method based on earthquake Krauklis waves. The urban karst water guide channel detection system comprises a piezoelectric type vibroseis, an array type earthquake receiving system and a multi-wave seismograph. The piezoelectric vibroseis is used for exciting seismic waves; and the multi-wave seismograph is used for acquiring seismic data acquired by each seismic acceleration sensor in the array type seismic receiving system. The structure can ensure the precision of seismic wave data acquisition each time; separating and extracting Krauklis waves from the acquired seismic wave data by adopting an energy-frequency double-selection method; then, based on the frequency dispersion characteristic of Krauklis waves, a mapping model of crack opening and wave field response is established; and finally, determining the opening degree of the urban karst water guide channel according to the mapping model, and performing inversion to obtain the dynamic flow of the urban karst water guide channel. Through treatment in the mode, the opening degree and the dynamic flow of the urban karst water guiding channel can be monitored, and finally prevention and control over hidden geological disasters are effectively improved.

According to the micro-seismic source intelligent positioning method and the model training method and system, physical laws such as a seismic wave equation are integrated, physical loss and data loss are comprehensively considered, the adaptability of the model to a complex geological environment is improved, micro-seismic source positioning challenges in the complex geological environment can be effectively dealt with, and the positioning accuracy of the micro-seismic source in the complex geological environment is improved. Noise immunity and positioning precision are improved, and more reliable technical support is provided for safety monitoring of underground engineering.

The invention relates to the technical field of roadway surrounding rock danger identification, and discloses a roadway surrounding rock danger identification model construction method, which comprises the steps of collecting multi-modal data, and generating preprocessed data through synchronous calibration and denoising; extracting a seismic wave frequency domain and image texture features, and generating a multi-modal feature matrix; in combination with a geological prior clustering mining abnormal mode, generating a labeled sample data set; generating a danger identification model based on a transfer learning and feature fusion training network; and the edge deployment model performs real-time reasoning, and generates an early warning result through an adaptive algorithm. According to the method, the frequency domain features of the seismic fluctuation signals and the depth texture features of the surrounding rock images are fused, the multi-modal feature matrix is constructed, abnormal mode mining is carried out in combination with geological prior knowledge, and early weak abnormal signals such as hidden fault slippage or asymmetric microfracture extension which are difficult to find by a single monitoring means can be effectively recognized.

The invention discloses an earthquake and electromagnetic joint inversion analysis method, belongs to the technical field of geological exploration, and can solve the problem that accurate judgment on geothermal resources is affected due to the fact that an existing single exploration method is difficult to obtain fine structural information and attribute information of a thermal reservoir. The method comprises the following steps: S1, constructing a multi-attribute geological model of a target area according to geological survey data of the target area, and respectively performing seismic exploration and electromagnetic exploration on the target area according to exploration working parameters of the target area to obtain seismic wave observation data and apparent resistivity observation data; s2, according to the multi-attribute geologic model and the exploration working parameters, seismic wave forward modeling and electromagnetic wave forward modeling are carried out on the target area, and seismic wave forward modeling data and apparent resistivity forward modeling data are obtained; and S3, determining a target geologic model of the target area according to the seismic wave observation data, the seismic wave forward modeling data, the apparent resistivity observation data and the apparent resistivity forward modeling data. The method is used for geothermal resource exploration.

The invention provides an active and passive source metal mine seismic exploration method, which comprises the following steps of: acquiring geological exploration data of a target area, and analyzing and judging to obtain a metallogenic favorable part of the target area; obtaining seismic geologic features of the target area and distribution features of natural field sources; according to the acquired geological exploration data, the passive source seismometers are arranged in a grid mode in a target area to form a passive source detection grid; the active source seismographs are arranged in a cross shape at the favorable metallogenic part to form two active source detection lines; artificial source seismic waves are excited, active source seismic data of the active source seismograph and passive source seismic data of the passive source seismograph are collected, and after the artificial source seismic waves are excited, the passive source seismic data of the passive source seismograph are continuously collected. The problem of active and passive source seismic data fusion acquisition is solved, the exploration cost is reduced, and the working efficiency and the detection precision are improved.

A method for removing tube wave interference from optical fiber acoustic wave sensing seismic data, including: acquiring seismic wavefield data which contains a tube wave and is collected by an optical fiber acoustic wave sensing instrument; calculating a root-mean-square amplitude of the waveform data cut on the seismic trace as an amplitude normalization factor; performing normalization processing on the amplitude value; performing de-tail mean filtering processing on the normalized amplitude value along the travel time of the tube wave, to obtain a predicted amplitude value; performing tube wave interference removal processing on each seismic trace, and performing inverse normalization processing to obtain the seismic wavefield data without tube wave interference. The method effectively suppresses the tube wave interference in the optical fiber acoustic wave sensing seismic data. An apparatus for removing tube wave interference from optical fiber acoustic wave sensing seismic data, and a computer device are further provided.

The invention discloses a mine earthquake intelligent identification method based on multi-mode deep learning and signal processing, and belongs to the technical field of mine earthquake identification. Aiming at the problems that a traditional mine earthquake signal processing method cannot fully utilize spatial relevance and frequency domain characteristics of signals in a sensor network, the automation level is relatively low, the response speed is slow, and the emergency processing effect is influenced, seismic waveform data and multi-modal characteristic input are adopted, so that the emergency processing effect is influenced. Based on text semantic features generated by a large language model (LLMs), frequency domain features extracted by Fourier transform and time sequence features after obspy processing, a dynamic graph recognition model of a fusion graph neural network (GNN) and Transform is constructed; graph structure edge weights are dynamically updated through cross correlation between signals, and the structure change of complex environments such as mine roadways is self-adapted.

The invention discloses a coal seam passive source seismic survey system and method based on deep learning, and relates to the technical field of coal mining safety monitoring and early warning, and the method comprises the following steps: S1, collecting the passive source seismic data of a coal seam area; s2, deep fusion features are extracted; s3, constructing a semi-supervised self-attention prediction network model; s4, extracting coal seam structure prediction parameters; s5, the corrected coal seam structure parameters are extracted; s6, constructing a fast seismic wave forward model, and calculating and outputting an initial model of the coal seam underground structure; s7, constructing a Transform optimization network, and extracting a depth feature vector; and S8, constructing a coal seam structure inversion model, and training iteration by using the improved batch depth Bayesian optimization algorithm. The method overcomes the limitation of manual analysis and slow response in a traditional coal seam seismic survey method, and provides an efficient and accurate solution for safe production of a coal mine.

The invention belongs to the technical field of geophysical exploration in an extremely complex tunnel environment, and provides an air-ground and hole-tunnel multi-angle combined tunnel unfavorable geology advanced detection system and method.A first main control unit is used for establishing a unified time reference and controlling an air-ground capacitive induced polarization detection unit to collect air-ground electric field data under the unified time reference; the second main control unit is used for controlling the hole-tunnel induced electric field and natural electric field combined detection unit to collect underground induced polarization response signals and continuous natural electric field signals and controlling the tunnel seismic electric field detection unit to collect seismic electric field signals and seismic wave signals; and the second main control unit is used for receiving the air-ground electric field data sent by the first main control unit, uniformly integrating the air-ground electric field data with an underground induced polarization response signal, a continuous natural electric field signal, a seismic electric field signal and a seismic wave signal which are acquired by the second main control unit to obtain a multi-field data set, and performing interpretation and inversion to obtain a potential unfavorable geologic body identification result. And the disaster source detection precision and efficiency in the extremely complex tunnel environment are obviously improved.

The invention discloses a microseismic source positioning method, device and system, and a storage medium. The method comprises the following steps: dividing a microseismic data sample data set into a training set and a test set; according to the training set, constructing a microseism inversion subnet containing a Swin Transform encoder, and according to the training set, constructing a microseism inversion subnet containing a Swin Transform encoder; the micro-seismic forward modeling subnet realizes seismic wave field continuation by inputting a micro-seismic source position and a speed model and utilizing a recurrent neural network structure and a convolution operator, and establishes a forward modeling subnet based on a wave equation; constructing an inversion-forward modeling closed-loop neural network according to the micro-seismic inversion subnet and the forward modeling subnet based on the wave equation; and inputting the test set into the inversion-forward closed-loop neural network to carry out micro-seismic source positioning. By adopting the technical scheme of the invention, the limitations on physical constraint, feature modeling and anti-noise capability in the prior art are overcome.

The invention relates to the technical field of rock mass earthquake dynamic response, and discloses a layered rock mass earthquake dynamic response multi-scale monitoring system and method, and the system comprises a rock mass three-dimensional geological module, a simulation setting module and an earthquake simulation module. The method comprises the following steps: establishing a layered rock mass three-dimensional geologic model; the method comprises the following steps: identifying a dangerous rock mass structural surface, monitoring slope displacement, and installing sound emission monitoring equipment at a key position to form a three-dimensional monitoring network; a layered rock mass and a broken area in the layered rock mass three-dimensional geological model are divided into discrete element areas, and a continuous medium area is divided into finite difference areas; coupling the discrete element region and the finite difference region; inputting the seismic wave as a dynamic boundary condition into the finite difference region, transmitting a response to the discrete element region through a coupling interface, and simulating a rock stratum and rock mass dynamic fracture process; and the crack propagation particle size and the original stress propagation are transmitted to the cloud platform. The geometric morphology and the internal structure of the rock mass can be accurately reflected.

The invention discloses a boundary-preserving seismic wave impedance inversion method, device, equipment and medium, and relates to the technical field of geological prospecting. Historical seismic data of a target area and logging curve data including logging impedance parameters are acquired, and seismic wavelet data are extracted from the historical seismic data; carrying out data processing on the seismic wavelet data to obtain target seismic wavelet data, constructing an initial impedance parameter model according to the logging curve data, carrying out forward modeling on the initial impedance parameter model, constructing a composite objective function comprising a boundary preserving constraint term and a prior information constraint term, and carrying out the forward modeling on the initial impedance parameter model based on the composite objective function. And the initial impedance parameter model is updated to obtain the target impedance parameter model, and the output impedance parameter inversion curve is obtained through the target impedance parameter model, so that the inversion precision and efficiency are improved, and the inversion stability and the geological reduction capability are improved.

The invention discloses a fault co-seismic sliding surface inversion method and system based on Beidou and artificial intelligence. The method comprises the steps of collecting Beidou data and seismic waveform data of a target area in real time, and performing preprocessing to extract co-seismic displacement and waveform features; fusing the Beidou deformation features and the seismic wave features to form a multi-modal feature vector; a fault prediction model is constructed based on a graph neural network, forward modeling and historical earthquake example data training are utilized, and multi-modal features are input to obtain preliminary fault sliding distribution prediction; rapid optimization under physical constraint is carried out through the elastic dislocation model, and a final fault sliding model conforming to the geophysical law is obtained; finally, uncertainty quantification is carried out, and a visual product of the dynamic process including sliding distribution, seismic moments and fault deformation is generated. According to the system, full-process automatic processing is achieved through cooperation of all the units, the inversion speed, precision and physical credibility are improved, and reliable support is provided for earthquake emergency response and disaster assessment.

The invention provides a tight gas micro-crack detection method and system based on seismic data, and relates to the technical field of tight gas micro-crack detection, and the method comprises the steps: obtaining seismic wave data, and obtaining the preprocessed seismic wave data through wavelet transform denoising, compressed sensing reconstruction and median filtering smoothing processing. Using a preset multi-dimensional feature extraction model to extract micro-crack related features from the preprocessed data, including crack edge, curvature, frequency and similarity features, and constructing a multi-dimensional feature matrix; and based on the feature matrix, modeling and training are carried out by adopting a convolutional neural network model of a spatial attention mechanism added with a channel, and then the crack probability of each to-be-detected position is predicted. And carrying out smoothing, denoising and connectivity analysis on the crack probability image through image conversion and optimization processing to obtain a final crack detection result. According to the invention, the precision of micro-crack detection can be effectively improved.

The invention provides a three-dimensional interface inversion method fusing geological constraints and a level set function, and relates to the technical field of geophysical inversion and imaging, and the method comprises the steps: obtaining geological prior information, carrying out the parametric modeling of an underground structure of a research region through a Gaussian random field, and generating a model parameter vector; and simulating the propagation process of seismic waves in a seismic wave speed model by using a forward modeling calculation method, carrying out iterative optimization on model parameter vectors based on an EKI set Kalman inversion method, recovering an underground three-dimensional structure in a research area, and carrying out interpretation and verification in combination with drilling data to obtain structural characteristics and resource distribution of the research area. According to the method, unified parametric modeling is performed on speed disturbance and interface forms through a Gaussian random field, and a level set function is introduced to represent discontinuous layering, fault cutting, inclusion embedding and other complex geological interfaces, so that collaborative inversion of a speed field and a structure boundary under the same optimization framework is realized; and geometric expressive power and geological consistency of the model are improved.