25 results about "Geophysical inversion" patented technology

The invention discloses an electromagnetic detection data efficient three-dimensional inversion method based on operator learning, and belongs to the technical field of geophysical inversion, and the method comprises the steps: constructing an operator learning data set which comprises a plurality of electromagnetic forward modeling models and forward modeling responses of the electromagnetic forward modeling models; performing aviation electromagnetic frequency domain forward calculation training based on the operator learning data set by using a DeepONet network to obtain a trained aviation electromagnetic three-dimensional forward operator; inputting the initial uniform half-space model, replacing a generator with an aviation electromagnetic three-dimensional positive operator, and constructing a WGAN network by taking a KAN network as a discriminator to carry out aviation electromagnetic three-dimensional inversion; training is carried out based on the loss function and the data fitting difference convergence condition to update the KAN network and the uniform half-space model, and finally the updated uniform half-space model is obtained to serve as an inversion result. According to the method, more efficient and accurate three-dimensional aviation electromagnetic inversion can be realized, and the method is more suitable for geophysical inversion application under complex data distribution.

The invention discloses a DeepLabV < 3 + >-based magnetotelluric two-dimensional inversion method, and belongs to the field of geophysical inversion methods, and the method comprises the steps: constructing a plurality of underground resistivity distribution models, and obtaining corresponding apparent resistivity and phase data through forward modeling calculation; designing a deep learning network based on a DeepLabV < 3 + > architecture, and enhancing the capability of capturing multi-scale geological features by using a cavity convolution and improved cavity spatial pyramid pooling (ASPP) module of the deep learning network; di ce Loss is adopted to replace a traditional cross entropy loss function, and the method is more suitable for the characteristics of anomaly and background information imbalance in geophysical observation data; according to the method, the network is trained and verified through the training sample set, and the optimal network parameters are obtained; the trained network is utilized to directly map the apparent resistivity and phase data of a test set into underground resistivity distribution, the end-to-end inversion process is realized, and the multi-scale feature extraction and semantic segmentation capabilities of the DeepLabV3 + network are utilized, so that compared with a traditional inversion method, the method has the advantages of being independent of an initial model, capable of obtaining a global optimal solution and the like.

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.

The invention provides a method for constructing a geoelectric structure prediction model based on time-frequency electromagnetism, and belongs to the field of oil exploration. The method comprises the following steps: firstly, establishing an initial reservoir model based on a model-driven geophysical inversion process; then geophysical data and a corresponding reservoir model data set are collected, classified and standardized, and construction of the data set is completed; establishing a reservoir geoelectric structure prediction model based on the deep learning model; and finally training the reservoir geoelectric structure prediction model by using the data set to obtain an optimized reservoir geoelectric structure prediction model. Wherein the reservoir comprises an ultra-deep carbonate rock reservoir; the invention further provides a method for constructing the model for predicting the geoelectric structure of the ultra-deep carbonate reservoir. Compared with the prior art, according to the geoelectric structure prediction model constructed by the method, when observation data is input, an inversion result can be obtained without executing an iteration process, and the reservoir oil-water relationship can be further analyzed according to the inversion result.

The invention belongs to the technical field of geophysical inversion, and relates to a gravity anomaly three-dimensional inversion method based on multi-scale feature fusion, which comprises the following steps: randomly generating density anomaly bodies with different numbers, densities, sizes and shapes to obtain an underground medium three-dimensional density model; carrying out forward modeling on the three-dimensional density model of the underground medium to obtain two-dimensional gravity abnormal data of the earth surface; a multi-scale feature fusion model is trained with the underground medium three-dimensional density model as output and the two-dimensional gravity anomaly data as input, and the multi-scale feature fusion model comprises a two-dimensional encoder, a three-dimensional decoder, a dimension converter and a multi-scale feature fusion device, wherein the dimension converter and the multi-scale feature fusion device are located between the two-dimensional encoder and the three-dimensional decoder; and inputting actually measured two-dimensional gravity anomaly data into the trained multi-scale feature fusion model, and performing inversion to obtain a three-dimensional density body constrained by a physical range. According to the method, the expression capability of the multi-scale features of the underground structure is enhanced, and the multi-solution problem and the deep resolution reduction problem of inversion can be relieved.

The invention discloses a slice quantitative analysis method for a target body positioning error in a geophysical inversion result, and the method comprises the following steps: building a parameter model, carrying out the sampling initialization, and carrying out the combined sampling of a position parameter and a model parameter of a target body through combining geological prior information. An error evaluation mechanism is introduced in the sampling process, and sampling results are guided to be concentrated towards areas with small errors. A plurality of collected position samples are used for constructing a three-dimensional uncertainty boundary structure, and geometric slicing is performed in a plurality of section directions to generate a series of error section images. The images are further subjected to weighted fusion according to the importance of different sections, and finally a three-dimensional positioning error thermodynamic diagram is generated and used for reflecting the uncertainty spatial distribution of the underground target position. According to the method, the spatial deviation degree in the inversion result can be expressed in a visual form, and a quantitative basis is provided for geological interpretation and resource evaluation.

The present application relates to the field of geophysical inversion and oil and gas reservoir prediction, and particularly relates to a method for predicting fractures of lacustrine carbonate rock based on fractional order total variation. The method comprises the following steps: obtaining an initial model of a to-be-inverted parameter by preprocessing seismic data; constructing a forward model of prestack elastic parameters and an objective function based on fractional order total variation; updating the inversion of the log matrix L by combining the alternating direction multiplier method, the gradient being zero and the extreme value, and eigenvalue decomposition; obtaining Young's modulus, Poisson's ratio and density according to the relationship between Young's modulus, Poisson's ratio and the elastic three-parameter log matrix L i+1 ; and obtaining the fracture prediction result by slicing the inversion result of Young's modulus and Poisson's ratio according to the horizon. The present application uses fractional order total variation to establish a sparse regularization term and construct an objective function of prestack elastic three-parameter seismic inversion, thereby improving the accuracy and stability of the inversion result and the accuracy of fracture prediction.

The application belongs to the technical field of geophysical inversion, and relates to a gravity anomaly three-dimensional inversion method based on multi-scale feature fusion, comprising the following steps: randomly generating density anomaly bodies with different number, density, size and shape to obtain a three-dimensional density model of underground medium; forward modeling the three-dimensional density model of underground medium to obtain two-dimensional gravity anomaly data of the ground surface; training a multi-scale feature fusion model with the three-dimensional density model of underground medium as output and the two-dimensional gravity anomaly data as input, wherein the multi-scale feature fusion model comprises a two-dimensional encoder and a three-dimensional decoder, and a dimension converter and a multi-scale feature fusion device located between the two-dimensional encoder and the three-dimensional decoder; inputting the measured two-dimensional gravity anomaly data into the trained multi-scale feature fusion model to obtain a three-dimensional density body constrained by a physical range through inversion. The application enhances the expression capability of multi-scale features of underground structures, and can relieve the multi-solution problem of inversion and the problem of resolution decline in deep parts.

The application discloses a regional lithology interpretation method, system, device and medium based on multiple physical characteristics, and the method comprises the following steps: acquiring regional logging data and geophysical data; preprocessing the regional logging data to obtain a target regional logging data set; establishing multiple classification models associated with the input of multiple physical characteristics; training and verifying the multiple classification models by using the target regional logging data set to determine the best classification model; performing multidimensional inversion on the geophysical data to obtain an inversion result; and inputting the inversion result into the best classification model for processing to obtain a lithology interpretation result. Therefore, by establishing multiple classification models associated with the input of multiple physical characteristics, and determining the model with the optimal classification effect, i.e. the best classification model, from the multiple classification models, the best classification model can be applied to the lithology interpretation of the geophysical inversion result, so that the lithology interpretation result with relatively better accuracy, efficiency and intelligence can be obtained.

The invention provides a goaf positioning method and system for different mining degrees, and belongs to the technical field of geophysical inversion, and the method comprises the steps: obtaining multi-temporal SAR image data and on-site mining degree data of a target coal mine area; constructing an integrated probability model suitable for different mining degrees; and setting a search range of goaf characteristic parameters and model parameters of the integrated probability model according to the field mining degree data, and optimizing the goaf characteristic parameters by using a nonlinear inversion algorithm to obtain the spatial-temporal position of the goaf. Technical support is provided for coal mining monitoring, illegal mining supervision and historical mining working face information in a large area.

The present invention discloses a pre-stack seismic direct inversion method for P-wave and S-wave velocity ratios, which is applied to the field of geophysical inversion. Currently, P-wave and S-wave velocity ratios are mostly calculated based on the P-wave and S-wave velocities obtained according to the AVO approximate formula, but the inversion accuracy is inevitably reduced. In response to the above problems, the present invention proposes a pre-stack seismic method for directly inverting the P-wave and S-wave velocity ratios, the method comprising: constructing a Bayesian nonlinear inversion algorithm based on the PSR-EI equation by deriving the elastic impedance (EI) equation of the P-wave and S-wave velocity ratio (PSR), and obtaining a robust P-wave and S-wave velocity ratio estimation by repeatedly iterating the objective function of the algorithm. The technology provided by the present invention can improve the inversion accuracy of the P-wave and S-wave velocity ratios, thereby improving the processing and interpretation efficiency and accuracy of actual work.

The application discloses a multi-channel inversion method and system based on an improved genetic algorithm, and relates to the technical field of geophysical inversion.The application adopts the improved genetic algorithm to realize inversion of seismic records when inverting the seismic records, adjusts fitness values of the inversion process, arranges fitness function values of each individual in descending order, increases the fitness values of the individuals corresponding to the first quarter of the fitness function values by a first preset value, reduces the fitness function values of the individuals corresponding to the last quarter of the fitness function values by a second preset value, and the fitness values of the individuals corresponding to the middle quarter of the fitness function values remain unchanged; then, selection and heredity are carried out by using the new fitness function values; thus, the excellent individuals cannot be eliminated due to the small number, and the poor individuals can be more easily eliminated due to the reduced fitness values, so that the genetic algorithm can better play the optimization effect, and the genetic algorithm can converge more quickly.

The present invention discloses an efficient three-dimensional inversion method for electromagnetic detection data based on operator learning, belonging to the field of geophysical inversion technology. The method comprises: constructing an operator learning dataset comprising several electromagnetic forward models and their forward responses; performing airborne electromagnetic frequency-domain forward modeling training using a DeepONet network based on the operator learning dataset to obtain a trained airborne electromagnetic three-dimensional forward operator; inputting an initial uniform half-space model into a generator replaced by an airborne electromagnetic three-dimensional forward operator, and constructing a WGAN network using a KAN network as a discriminator to perform airborne electromagnetic three-dimensional inversion; training based on a loss function and a data fit error convergence condition to update the KAN network and the uniform half-space model, ultimately obtaining an updated uniform half-space model as the inversion result. The present invention can achieve more efficient and accurate three-dimensional airborne electromagnetic inversion and is more suitable for geophysical inversion applications with complex data distributions.

The present invention discloses a method for extracting adverse geological features from tunnel geophysical inversion data, comprising the following steps: pre-processing face data and advance drilling data to obtain a set of interpolation points of a target type of adverse geological body, performing interpolation calculations on the point set to obtain a voxel model of the adverse geological body; using advance drilling data to perform quality assessment on the interpolation results of the adverse geological body; using geophysical inversion data to construct a geophysical field voxel model, obtaining the intersection of the geophysical field voxel model with the adverse geological body voxel model to obtain an adverse geological feature voxel model; mapping the coordinates of the adverse geological feature voxel model to the geophysical inversion data, extracting a local coordinate set and resampling it according to the corresponding accuracy to obtain a set of coordinate points of the adverse geological feature. The present invention can integrate the adverse geological prior information revealed by the face data and the advance drilling data, accurately extract the adverse geological features in the multimodal geophysical inversion data, and facilitate the clarification of the geophysical response characteristics of the adverse geological body.

The application provides a kind of side slope fissure passive source seismic wave imaging method, the patent adopts multi-wave multi-component seismograph, simultaneously gathers multiple seismic wave signals on three components, including surface wave, body wave and the like.This design significantly improves the richness of data, so that in data processing, a variety of seismic wave signals can be comprehensively utilized for inversion analysis.By joint processing and constraint of different wave types, the multi-solution problem in geophysical inversion can be effectively reduced, and the stability and reliability of the results are improved.In addition, the acquisition of multi-component data makes the imaging more comprehensive, especially in the characterization of fissure directionality and geometric features, which has obvious advantages.

This invention discloses a natural electric field source inversion method based on a genetic algorithm, belonging to the field of geophysical inversion technology. It solves the technical problems of low inversion accuracy and large source location errors in natural electric field inversion caused by optimization algorithms. The key points of the technical solution are: first, expanding the input electrical resistivity data using head-to-tail cross-interpolation; second, using the extreme values ​​and dispersion of the electrical resistivity data as the center and radius of the population distribution area; third, forming a population by describing individuals based on the number and location of source sources; and finally, using a genetic algorithm based on double-threshold cross-iteration to achieve natural electric field source inversion with smaller source fitting errors. This application can effectively utilize a small amount of input electrical resistivity data, reasonably expand and extract extreme value features, and largely solve the problems of large inversion fluctuations caused by large spatial areas, small data volumes, and the inherent multiple solutions in geophysical inversion, while achieving smaller source fitting errors.

The invention discloses a salt cavern hydrogen storage reservoir group geologic model construction method based on multi-means technology fusion. The method comprises the steps that a preliminary model is obtained through electromagnetic and earthquake independent inversion; performing primary joint inversion based on cross gradient constraint, and constructing a macroscopically consistent model pair; constructing a multi-physics field database through drilling sampling; and finally, by using drilling data as strong prior information, constructing a well-seismic-electric constraint term containing a rock physical statistical mapping relation and a formation interface geometric trend, carrying out constrained secondary inversion, and outputting a final three-dimensional geologic model. According to the technical scheme, through combination of two-stage inversion and strong geometric constraint, the problem of multiplicity of solutions of geophysical inversion is solved, one-dimensional well hole actual measurement data is expanded into three-dimensional space constraint by using a strong constraint term, the reliability of a geologic model in a well-free area is improved, and the recognition capability of a hydrogen storage bank group stratum boundary is enhanced.

This invention relates to the field of geophysical inversion technology and provides a high-precision joint gravity and magnetic property vector inversion method based on data transformation. The method includes the following steps: Step S1, dividing the subsurface space into N block units in spherical coordinates; Step S2, establishing a kernel function matrix A; Step S3, establishing a regularized inversion objective function; Step S4, constructing Gram constraint terms; Step S5, establishing the objective function for Gram joint property vector inversion; Step S6, iteratively calculating the density and magnetic vector results using the objective function. This invention enhances the correlation between different parameters of the Gram constraint based on data transformation, and utilizes the strengthened constraint relationship for joint gravity and magnetic property vector inversion, thus improving the resolution of the inversion results.

The application discloses a slice quantitative analysis method for positioning error of a target body in geophysical inversion results, and comprises the following steps: a parameter model is established and sampling initialization is performed; position parameters of the target body and model parameters are jointly sampled by combining with geological prior information. An error evaluation mechanism is introduced in the sampling process to guide the sampling results to concentrate in the area with smaller error. The collected multiple position samples are used to construct a three-dimensional uncertainty boundary structure, and a series of error cross-section images are generated by performing geometric slicing in multiple profile directions. Further, the images are weighted and fused according to the importance of different profiles, and finally a three-dimensional positioning error heat map is generated to reflect the uncertainty spatial distribution of the underground target position. The application can express the spatial deviation degree in the inversion results in an intuitive form, and provides a quantitative basis for geological interpretation and resource assessment.

The invention provides a cross-band rock physical parameter prediction method under a pressure condition. According to the method, unified and accurate prediction of key parameters such as longitudinal wave velocity, transverse wave velocity, Young modulus, Poisson's ratio and the like under a full-band and multi-pressure condition from a static state, namely a low frequency, to a dynamic state, namely a high frequency is realized by utilizing strong nonlinear mapping and feature correlation capturing capability of a deep neural network and a Transform architecture. The method effectively overcomes the limitation of traditional experiments and models, and provides complete and reliable rock physical parameter support for geophysical inversion and engineering design.

The present application is suitable for the field of geophysical inversion technology, and provides a non-structure triangular grid gravity inversion method based on equivalent coplanar-function fitting, comprising the following steps: step S1, obtaining the Bouguer gravity anomaly; step S2, carrying out tetrahedron dissection on the underground space according to the terrain data; step S3, calculating the kernel function according to the equivalent principle; step S4, carrying out iterative inversion according to the function fitting method to obtain the density distribution of the underground space. The present application accelerates the kernel function calculation and iterative calculation process by combining the equivalent principle and the function fitting method, effectively improves the calculation efficiency, saves the calculation time, effectively solves the high-efficiency calculation problem of gravity inversion under the non-structure grid dissection of large data in the undulating terrain area, and is suitable for mineral exploration with high resolution in the undulating terrain area and scientific research on large-area deep complex structure.

The invention relates to the technical field of fractured thermal reservoir identification, in particular to a complex fractured thermal reservoir joint identification method and device, and the method specifically comprises the following steps: 1, obtaining and collection of multi-source data: the obtaining and collection of the multi-source data comprise the collection of geological data, geophysical data and logging data; step 2, data preprocessing, wherein the data preprocessing comprises de-noising processing of seismic data and normalization processing of logging data; step 3, establishing a geologic model; step 4, carrying out geophysical inversion; 5, performing logging data feature extraction; step 6, establishing different fracture type probability models for fracture type thermal reservoirs involving different lithology; step 7, fracture connectivity analysis is carried out; 8, heat storage parameter calculation is carried out; step 9, carrying out model verification; and step 10, result visualization processing is carried out. According to the invention, the problems of limited geophysical logging information acquisition range and low seismic exploration resolution are solved.

The invention belongs to the field of geophysical inversion, and particularly discloses a shear wave velocity two-dimensional profile probability inversion method based on multi-source data fusion, and the method comprises the steps: introducing transverse constraint and drilling information, and achieving the joint probability inversion of a multi-measurement-point shear wave velocity profile; under a Bayesian framework, uniformly regarding multi-measurement-point frequency dispersion curve data, transverse constraint and drilling data as observation data items, establishing posterior probability distribution of multi-measurement-point shear wave velocity profile model parameters under three types of data constraints, and introducing various types of observation information step by step by adopting a sequential Bayesian updating strategy; and finally solving the Bayesian equation of each stage to obtain the posterior distribution of the shear wave velocity profile of the multiple measuring points. The method of the invention can effectively quantify the identification uncertainty of the shear wave velocity profile of multiple measuring points, enhance the continuity and physical consistency of adjacent profiles in the transverse direction, and improve the identifiability of the shear wave velocity profile.

The application discloses a geophysical exploration inversion method based on ensemble learning and belongs to the technical field of engineering geophysical exploration inversion. In order to solve the problems that multi-source geophysical exploration inversion is strong in non-uniqueness, is dependent on experience for interpretation and is difficult to be uniformly constrained with in-situ drilling test, the application fuses features through a deep neural network, carries out posterior sampling after condition regularization flow and applies interface monotone order and parameter boundedness and drilling anchoring constraint, combines neural operator proxy forward and consistent screening iteration, and outputs interfaces, parameters and uncertainty by means of multi-member adaptive weighted ensemble, so that the technical effects of stable inversion and multi-solution credible screening are realized.

The application discloses a kind of based on low rank structure constraint multi-geophysical field joint inversion method, it is related to geophysical field joint inversion field, constructs low rank structure coupling joint inversion mathematical optimization model, utilizes singular value and rank to represent physical property parameter space structure, establishes including data fitting term, model constraint term and low rank structure constraint target function, adopts Schatten-p norm to carry out model multi-property low rank constraint, obtains optimization model optimal solution using nonlinear optimization method, establishes the multi-property attribute model with low rank physical property structure, not only can effectively represent and evaluate the global correlation of multi-physical property model in physical property space, but also can reduce the rank of multi-physical property in physical property space by reduction, gradually enhance the model multi-property correlation in iterative calculation process, reduce geophysical inversion multi-solution and interpretation uncertainty, finally establish the multi-property attribute model with structure similarity.