A method for analyzing seismic exploration well depth based on complex ground surface environment

Abstract

The application relates to the technical field of geological exploration, and particularly discloses a seismic exploration well depth analysis method based on a complex surface environment, which is used to solve the problems of blindness and low efficiency of well depth setting in a traditional manual experience method, and comprises the following steps: analyzing complex surface and shallow surface geological conditions, constructing a middle-deep layer geological model, designing a zoned well depth test, collecting multi-gradient well depth test data, automatically analyzing seismic reflection data based on deep learning, intelligently optimizing inversion of well depth, verifying and constructing a three-dimensional data body and geology application analysis of optimized well depth on site; and through a dynamic model of real-time feedback, the well depth gradient and geomorphology identification parameters are dynamically optimized for different geomorphic regions such as bedrock exposed areas, farmland sections, weathered material areas and valley areas, so that automatic adaptive adjustment of the seismic exploration well depth is realized.

Description

Technical Field

[0001] This invention relates to the field of seismic exploration technology, and more specifically, to a method for seismic exploration well depth analysis based on complex surface environments. Background Technology

[0002] In the publicly available CNKI paper "Design of 3D Seismic Exploration Acquisition Parameters - A Case Study of Changping Coalfield in Southern Qinshui Basin," the author, using the Qinshui Basin coalbed methane field as an example, analyzed well depth and charge parameters using an 8-line, 12-shot bundle-shaped 3D observation system in bedrock exposed areas, farmland areas, weathered areas, and valley areas. By analyzing the imaging effects, the author obtained the best-performing well depth and charge in different near-surface environments through excitation depth tests, charge tests, receiver condition tests, and optimal observation range tests under various conditions. This facilitates the construction of 3D seismic profiles with distinct wave group characteristics, good phase axis continuity, high energy, rich geological information, and clear shallow and deep seismic interface reflection waves for exploration areas with similar geological conditions. Seismic time profiles are used to facilitate the construction of a three-dimensional data volume for the region, laying the foundation for subsequent seismic sequence lattice, geological inversion, and prediction. However, relying solely on experience to evaluate imaging effects from various test points in different experiments to obtain excitation well depths for different environments is inefficient, leads to high personnel fatigue, and has poor accuracy. This is especially true in areas with complex terrain, large relative elevation differences, dramatic lateral lithological variations, and Quaternary loose areas, where seismic reflection wave energy absorption and attenuation are severe, seismic wavefields are complex, and wavelet consistency is poor. Furthermore, the target layers are numerous and their occurrence conditions vary greatly. How to extract effective features from massive amounts of seismic data, experimental data, and image data to refine these data and analyze suitable seismic exploration well depths has become a major challenge in the acquisition phase. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of existing technologies, this invention provides a seismic exploration well depth analysis method based on complex surface environments. Through a real-time feedback dynamic model, it dynamically optimizes well depth gradients and geomorphological identification parameters for different geomorphic areas such as bedrock exposed areas, farmland areas, weathered areas, and valley areas, thereby achieving automatic adaptive adjustment of seismic exploration well depth.

[0004] In this technical solution, the shallow strata are strata ranging from 5m to 200m, and the medium-deep strata are strata ranging from 200m to 1000m.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A seismic exploration well depth analysis method based on complex surface environments includes the following steps:

[0007] Step 1: Analyze complex surface and shallow geological conditions: Obtain surface orthophotos and digital elevation models through aerial photography to identify landform features. Use high-frequency surface radar to detect the thickness and distribution of loose structures in the shallow layer. Through on-site sampling and laboratory analysis, clarify the distribution of soil and rock particles and the degree of loose compaction. Analyze complex surface and shallow geological conditions. Dynamic model-driven fine image segmentation and real-time parameter adjustment improve the accuracy of surface and shallow geological analysis.

[0008] Step 2: Construct a medium-deep geological model: Determine the burial depth, thickness, and surrounding lithology of the target strata by integrating historical borehole data. Use inter-well logging data to determine the characteristics of the T3 wave reflection interface and its shielding effect on the coal seam. Establish a medium-deep seismic geological model. Combine borehole and logging data for precise quantitative modeling to achieve accurate analysis of the T3 wave shielding effect.

[0009] Step 3, Designing Zoned Well Depth Tests: Based on the analysis results of Step 1 and Step 2, the region is divided using a geographic information system. Several different levels of well depth gradients are designed for different seismic wavefield characteristics. Real-time dynamic model feedback and fine-tuning of the stepped well depth gradient are integrated to achieve high precision and dynamic adaptation of well depth zoning.

[0010] Step 4: Collect multi-gradient well depth test data: Conduct seismic excitation tests according to well depth gradients in each test zone, and collect shallow and medium-deep seismic reflection data using nodal three-dimensional seismic sensors;

[0011] Step 5, Automatic Seismic Reflection Data Analysis Based on Deep Learning: Construct an automatic identification system based on convolutional neural networks, train the model using labeled training data, automatically extract waveform continuity, signal-to-noise ratio, and attenuation rate using the trained automatic identification system, score the data, and optimize the model using transfer learning.

[0012] Step 6, Well Depth Intelligent Optimization Inversion: Establish a machine learning mapping model with the model evaluation results from Step 5 and the corresponding well depth data, and use the XGBoost intelligent optimization algorithm to perform well depth inversion analysis to determine the optimal well depth parameters for each partition.

[0013] As a further aspect of the present invention, the method further includes:

[0014] Step 7, Field verification of optimized well depth and construction of 3D data volume: Use the optimized well depth data to conduct actual field tests, re-acquire data, use data processing software to establish a 3D seismic time profile, analyze wavefield characteristics and wavelet consistency, use 3D seismic data to refine wavefield imaging, and clarify the characteristics of shallow and medium-deep layers.

[0015] Step 8, Geological Application Analysis: Based on the optimized 3D data volume, conduct detailed seismic sequence grid construction and geological inversion analysis to clarify the precise distribution and occurrence conditions of coal seams, analyze geological structural changes and coalbed methane occurrence characteristics, and provide a geological structural interpretation report.

[0016] As a further aspect of the present invention, in step one, the specific method for identifying landform features by acquiring orthophotos and digital elevation models of the landform through aerial photography is to use supervised classification to obtain input feature data from the images and identify various landform types, including:

[0017] Image regions with RGB brightness greater than 200, color standard deviation less than 20, texture contrast greater than 0.7 and texture entropy less than 3 are identified as bedrock exposed areas.

[0018] Image regions with an RGB color standard deviation of less than 30, a regularity coefficient of greater than 0.8, and a boundary sharpness index of greater than 0.75 are identified as farmland areas.

[0019] Image regions with an RGB color standard deviation greater than 40, a crack density greater than 5 lines / m², and a texture entropy greater than 5 are identified as weathered areas.

[0020] Image regions with a DEM slope greater than 15°, a terrain depth difference greater than 2 meters, and a valley width standard deviation less than 0.5m are identified as valley areas.

[0021] As a further aspect of the present invention, in step one:

[0022] The input feature data of the bedrock exposed area is dynamically adjusted by a regional historical geological feature learning model, with RGB brightness ±10, color standard deviation ±5, texture contrast ±0.1, and texture entropy ±0.5.

[0023] The input feature data of farmland plots are adaptively adjusted using an agricultural cycle variation model, with color standard deviation ±5, regularity coefficient ±0.1, and boundary sharpness index ±0.05.

[0024] The input feature data of the weathered area are dynamically adjusted by the weathering process dynamic monitoring model, with color standard deviation ±5, crack density ±1 line / m², and texture entropy ±0.5.

[0025] In the valley area, the slope ±2°, depth difference ±0.5m, and valley width standard deviation ±0.1m are dynamically adjusted using a geomorphological evolution trend analysis model.

[0026] As a further aspect of the present invention, the regional historical geological feature learning model is a machine learning model constructed using the regional stability index and historical rock weathering rate.

[0027] The agricultural cycle variation model is a time series analysis model established based on seasonal crop planting patterns, crop types, and growth cycles.

[0028] The dynamic monitoring model for the weathering process is a numerical model based on historical data analysis, which is based on real-time rainfall, temperature, and historical weathering rate.

[0029] The geomorphic evolution trend analysis model is a real-time analysis model built based on historical erosion, rainfall data, and geomorphic evolution trends.

[0030] As a further aspect of the present invention, in step one, the image processing steps for acquiring high-precision orthophotos and digital elevation models of the land surface using drone aerial photography, identifying landform features, and achieving preliminary area segmentation are as follows:

[0031] Step 11: Perform radiometric and orthorectification corrections on the acquired aerial image data;

[0032] Step 12: Using the Mean-shift algorithm, based on color similarity and spatial proximity, the original image is segmented into several homogeneous region patches;

[0033] Step 13: Extract the image quantization features of each segmented region and initially establish an image feature library for different landform categories;

[0034] Step 14: Classify the initially segmented regions using the random forest classification algorithm, and preliminarily identify the types of various landform regions based on the classification results;

[0035] Step 15: Adaptive image refinement processing is performed on the initially identified areas: A machine learning model is constructed based on historical geological stability index and rock weathering rate. The RGB brightness, color standard deviation, texture contrast, and texture entropy thresholds are dynamically adjusted. The thresholds are optimized based on feedback results from regional features. The segmentation boundaries of exposed bedrock areas are finely adjusted. A time-series analysis model of seasonal agricultural cycle changes is constructed. The color standard deviation, regularity coefficient, and boundary sharpness index are dynamically adjusted. Farmland boundaries and crop types are finely segmented and identified. Based on regional historical weathering rate and real-time climate monitoring data, a dynamic data value analysis model is constructed. The fissure identification density and texture entropy index are dynamically adjusted. The contour boundaries of weathered areas are finely adjusted. Combined with digital elevation data, a geomorphic evolution trend analysis model is constructed. The slope, geomorphic depth difference, and valley width thresholds are adjusted in real time to finely depict the contour details of valley areas.

[0036] As a further aspect of the present invention, in step one, a high-frequency surface radar is used to detect the thickness and spatial distribution characteristics of the shallow loose structure. The acquired characteristic data includes the loose layer thickness, the intensity value of the reflected wave at the loose layer interface, and the distribution of radar wave propagation velocity. Through on-site sampling and laboratory analysis, the distribution of soil particles and the degree of loose compaction are determined. The acquired data includes the particle size distribution, gradation coefficient, penetration resistance value, and permeability coefficient of soil particles.

[0037] As a further aspect of the present invention, in step two, the process of constructing the medium-deep seismic geological condition model includes:

[0038] Step 21, Historical Drilling Data Analysis: Based on existing historical drilling data, quantitatively determine the target stratum depth, coal seam thickness, surrounding rock type and lithological physical properties, including density, porosity and acoustic velocity.

[0039] Step 22: Determine T3 wave characteristics using inter-well logging methods: Use inter-well logging methods such as acoustic logging, density logging, and resistivity logging to determine the acoustic velocity, density, and resistivity of the T3 wave reflection interface. Analyze the intensity of the T3 wave reflected wave, the wave continuity index, and its attenuation ratio of the seismic reflected wave energy of the underlying coal seam to form a shielding effect intensity map, and quantitatively analyze the target formation characteristics and the T3 wave shielding effect.

[0040] As a further aspect of the present invention, in step three, the method for well depth testing zoning based on the data and features obtained in steps one and two includes:

[0041] Step 31: Import the acquired data into the GIS system: Import the data obtained in Step 1, such as the thickness of the loose layer, the intensity of the interface reflected wave, the radar wave propagation speed, the particle distribution, the penetration resistance, and the data obtained in Step 2, such as the burial depth of the medium-deep coal seam, the coal seam thickness, the type of surrounding rock and its lithological physical properties, and the intensity of the T3 wave shielding effect, into the GIS system.

[0042] Step 32, Design test well depth zoning: In the GIS environment, spatial interpolation is used to perform spatial continuity analysis on the imported data to generate a continuous spatial distribution layer. The overlay analysis function of the GIS system is used to perform overlay analysis on the shallow and medium-deep key parameter layers, and regional analysis is performed based on parameter similarity.

[0043] Step 33: Set the well depth gradient for each zone: Combine the data on the thickness of the shallow loose structure and the burial depth of the medium and deep coal seams within the zone to determine the appropriate well depth gradient for each zone, and output a well depth test zone map with clear markings under different surface and geological conditions in the GIS system platform.

[0044] As a further aspect of the present invention, in step 33, the specific method for determining each suitable well depth gradient, based on data on various topographic zones, the thickness of shallow loose structures within a zone, and the burial depth of medium-deep coal seams within a zone, includes:

[0045] Bedrock exposed area: The initial well depth gradient is 5m-10m, and the gradient range is dynamically adjusted by ±1m based on the feedback from the real-time updated rock weathering rate model;

[0046] Farmland areas: The initial well depth gradient is set to 10m-15m, and the gradient range is dynamically adjusted by ±1.5m based on feedback from the real-time agricultural cycle change model.

[0047] Weathered material zone: The initial well depth gradient is set to 15m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time weathering process dynamic monitoring model;

[0048] In valley areas: the initial well depth gradient is set to 10m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time geomorphological evolution trend analysis model.

[0049] The technical advantages of this invention, a seismic exploration well depth analysis method based on complex surface environments, are as follows: This invention comprehensively utilizes multi-source geological data and a dynamic model-driven mechanism with real-time feedback to accurately adapt to the differentiated characteristics of different landforms in bedrock exposed areas, farmland areas, weathered areas, and valley areas. It achieves refined and automated dynamic adjustment of well depth gradient parameters, effectively improving the accuracy, reliability, and analysis efficiency of well depth parameter determination in complex surface environments. It effectively overcomes the problems of low efficiency, high fatigue, and insufficient accuracy of traditional manual experience-based judgment processes, and improves the quality of seismic wavefield imaging in complex terrain areas. This provides a reliable and high-quality data foundation for subsequent refined seismic sequence framework construction, geological inversion analysis, and coalbed methane exploration prediction. Attached Figure Description

[0050] Figure 1 This is a flowchart of a seismic exploration well depth analysis method based on complex surface environments according to the present invention;

[0051] Figure 2 This is a flowchart illustrating the specific image processing operations for identifying landform features and achieving preliminary region segmentation in this invention.

[0052] Figure 3 This is a diagram illustrating the construction process of the deep seismic geological condition model in this invention;

[0053] Figure 4 This is a flowchart of the method for well depth testing and zoning based on the data and features obtained in steps one and two according to the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] like Figure 1 As shown, the present invention proposes a seismic exploration well depth analysis method based on complex surface environments, comprising the following steps:

[0056] Step 1: Analyze complex surface and shallow geological conditions: Obtain surface orthophotos and digital elevation models through aerial photography to identify landform features. Use high-frequency surface radar to detect the thickness and distribution of loose structures in the shallow layer. Through on-site sampling and laboratory analysis, clarify the distribution of soil and rock particles and the degree of loose compaction. Analyze complex surface and shallow geological conditions. Dynamic model-driven fine image segmentation and real-time parameter adjustment improve the accuracy of surface and shallow geological analysis.

[0057] Step 2: Construct a medium-deep geological model: Determine the burial depth, thickness, and surrounding lithology of the target strata by integrating historical borehole data. Use inter-well logging data to determine the characteristics of the T3 wave reflection interface and its shielding effect on the coal seam. Establish a medium-deep seismic geological model. Combine borehole and logging data for precise quantitative modeling to achieve accurate analysis of the T3 wave shielding effect.

[0058] Step 3, Designing Zoned Well Depth Tests: Based on the analysis results of Step 1 and Step 2, the region is divided using a geographic information system. Several different levels of well depth gradients are designed for different seismic wavefield characteristics. Real-time dynamic model feedback and fine-tuning of the stepped well depth gradient are integrated to achieve high precision and dynamic adaptation of well depth zoning.

[0059] Step 4: Collect multi-gradient well depth test data: Conduct seismic excitation tests according to well depth gradients in each test zone, and collect shallow and medium-deep seismic reflection data using nodal three-dimensional seismic sensors;

[0060] Step 5, Automatic Seismic Reflection Data Analysis Based on Deep Learning: Construct an automatic identification system based on convolutional neural networks, train the model using labeled training data, automatically extract waveform continuity, signal-to-noise ratio, and attenuation rate using the trained automatic identification system, score the data, and optimize the model using transfer learning.

[0061] Step 6, Well Depth Intelligent Optimization Inversion: Establish a machine learning mapping model with the model evaluation results from Step 5 and the corresponding well depth data, and use the XGBoost intelligent optimization algorithm to perform well depth inversion analysis to determine the optimal well depth parameters for each partition.

[0062] It should be noted that the method further includes:

[0063] Step 7, Field verification of optimized well depth and construction of 3D data volume: Use the optimized well depth data to conduct actual field tests, re-acquire data, use data processing software to establish a 3D seismic time profile, analyze wavefield characteristics and wavelet consistency, use 3D seismic data to refine wavefield imaging, and clarify the characteristics of shallow and medium-deep layers.

[0064] Step 8, Geological Application Analysis: Based on the optimized 3D data volume, conduct detailed seismic sequence grid construction and geological inversion analysis to clarify the precise distribution and occurrence conditions of coal seams, analyze geological structural changes and coalbed methane occurrence characteristics, and provide a geological structural interpretation report.

[0065] It should be noted that in step one, the specific method for identifying landform features by acquiring orthophotos and digital elevation models of the landform through aerial photography is to use supervised classification to obtain input feature data from the images and identify various landform types, including:

[0066] Image regions with RGB brightness greater than 200, color standard deviation less than 20, texture contrast greater than 0.7 and texture entropy less than 3 are identified as bedrock exposed areas.

[0067] Image regions with an RGB color standard deviation of less than 30, a regularity coefficient of greater than 0.8, and a boundary sharpness index of greater than 0.75 are identified as farmland areas.

[0068] Image regions with an RGB color standard deviation greater than 40, a crack density greater than 5 lines / m², and a texture entropy greater than 5 are identified as weathered areas.

[0069] Image regions with a DEM slope greater than 15°, a terrain depth difference greater than 2 meters, and a valley width standard deviation less than 0.5m are identified as valley areas.

[0070] Supervised classification is used to perform preliminary zoning of aerial orthophotos and DEM data. By utilizing RGB brightness, color standard deviation, texture contrast, texture entropy, slope, geomorphic depth difference, and valley width standard deviation, bedrock exposed areas, farmland areas, weathered areas, and valley areas are accurately identified. Combined with dynamic adjustment model feedback, the segmentation boundaries are refined in real time, which effectively improves the adaptability of geomorphic zoning and seismic well depth, and realizes the fine zoning and intelligent optimization of well depth gradient under complex surface conditions.

[0071] It should be noted that in step one:

[0072] The input feature data of the bedrock exposed area is dynamically adjusted by a regional historical geological feature learning model, with RGB brightness ±10, color standard deviation ±5, texture contrast ±0.1, and texture entropy ±0.5.

[0073] The input feature data of farmland plots are adaptively adjusted using an agricultural cycle variation model, with color standard deviation ±5, regularity coefficient ±0.1, and boundary sharpness index ±0.05.

[0074] The input feature data of the weathered area are dynamically adjusted by the weathering process dynamic monitoring model, with color standard deviation ±5, crack density ±1 line / m², and texture entropy ±0.5.

[0075] In the valley area, the slope ±2°, depth difference ±0.5m, and valley width standard deviation ±0.1m are dynamically adjusted using a geomorphological evolution trend analysis model.

[0076] By introducing regional historical geological feature learning models, agricultural cycle change models, weathering process dynamic monitoring models, and geomorphic evolution trend analysis models, the thresholds of various image feature parameters in bedrock exposed areas, farmland areas, weathered areas, and valley areas are dynamically adjusted. This optimizes geomorphic recognition accuracy and contour details in real time, effectively enhancing zoning accuracy and robustness. It enables well depth analysis to accurately match geologically complex areas, significantly improving seismic imaging accuracy and geological information extraction efficiency, and ensuring the accuracy and automation of well depth optimization in complex environments.

[0077] It should be noted that the learning model of the regional historical geological characteristics is a machine learning model constructed using the regional stability index and the historical rock weathering rate.

[0078] The agricultural cycle variation model is a time series analysis model established based on seasonal crop planting patterns, crop types, and growth cycles.

[0079] The dynamic monitoring model for the weathering process is a numerical model based on historical data analysis, which is based on real-time rainfall, temperature, and historical weathering rate.

[0080] The geomorphic evolution trend analysis model is a real-time analysis model built based on historical erosion, rainfall data, and geomorphic evolution trends.

[0081] The regional historical geological feature learning model selects the regional stability index and historical rock weathering rate because they reflect the long-term stability and weathering rate of the regional rock mass, directly related to the degree of weathering and stability changes, and can dynamically and accurately distinguish the weathering differences between bedrock exposed areas and other areas. The agricultural cycle variation model selects seasonal crop planting patterns, crop types, and growth cycles. These parameters can accurately reflect the seasonal changes and regular characteristics of farmland, which is conducive to the adaptive adjustment of farmland area image features and ensures recognition accuracy. The weathering process dynamic monitoring model selects real-time rainfall, temperature, and historical weathering rate. These parameters can dynamically reflect the intensity and speed of weathering process changes and timely capture the changing characteristics of rock fissure development in weathered areas to adapt to the actual weathered area feature changes. The geomorphological evolution trend analysis model selects historical erosion degree, rainfall data, and geomorphological evolution trends because these data can dynamically reflect the continuous changing trends of slope, width, and erosion depth in valley areas, thereby determining the contour features of valley areas in real time and accurately.

[0082] It should be noted that, as Figure 2 As shown, in step one, the image processing steps for acquiring high-precision orthophotos and digital elevation models of the land surface using drone aerial photography, identifying landform features, and achieving preliminary area segmentation are as follows:

[0083] Step 11: Perform radiometric and orthorectification corrections on the acquired aerial image data;

[0084] Step 12: Using the Mean-shift algorithm, based on color similarity and spatial proximity, the original image is segmented into several homogeneous region patches;

[0085] Step 13: Extract the image quantization features of each segmented region and initially establish an image feature library for different landform categories;

[0086] Step 14: Classify the initially segmented regions using the random forest classification algorithm, and preliminarily identify the types of various landform regions based on the classification results;

[0087] Step 15: Adaptive image refinement processing is performed on the initially identified areas: A machine learning model is constructed based on historical geological stability index and rock weathering rate. The RGB brightness, color standard deviation, texture contrast, and texture entropy thresholds are dynamically adjusted. The thresholds are optimized based on feedback results from regional features. The segmentation boundaries of exposed bedrock areas are finely adjusted. A time-series analysis model of seasonal agricultural cycle changes is constructed. The color standard deviation, regularity coefficient, and boundary sharpness index are dynamically adjusted. Farmland boundaries and crop types are finely segmented and identified. Based on regional historical weathering rate and real-time climate monitoring data, a dynamic data value analysis model is constructed. The fissure identification density and texture entropy index are dynamically adjusted. The contour boundaries of weathered areas are finely adjusted. Combined with digital elevation data, a geomorphic evolution trend analysis model is constructed. The slope, geomorphic depth difference, and valley width thresholds are adjusted in real time to finely depict the contour details of valley areas.

[0088] This method uses high-precision aerial images and DEM data correction, Mean-shift algorithm for preliminary region segmentation, and random forest classification algorithm to achieve preliminary identification of landform types. It also utilizes historical geological stability index and rock weathering rate model, agricultural cycle variation time series analysis model, weathering process dynamic monitoring model, and landform evolution trend real-time analysis model to dynamically and adaptively finely adjust the image feature thresholds for bedrock exposure areas, farmland areas, weathered areas, and valley areas. This solves the technical problems of difficult manual identification, large subjective errors, and insufficient regional division accuracy caused by blurred landform boundaries and large variations in geological conditions in complex terrain areas. It significantly improves the accuracy and efficiency of landform boundary identification and fine characterization, ensuring accurate and real-time adaptability for subsequent seismic well depth parameter gradient design. Ultimately, it achieves automation, efficiency, and accuracy in determining seismic exploration well depth parameters in complex surface environments, laying a solid data foundation for high-quality seismic exploration imaging and subsequent fine geological interpretation.

[0089] It should be noted that in step one, high-frequency surface radar is used to detect the thickness and spatial distribution characteristics of the shallow loose structure. The characteristic data obtained include the thickness of the loose layer, the intensity value of the reflected wave at the interface of the loose layer, and the distribution of radar wave propagation velocity. Through field sampling and laboratory analysis, the distribution of soil particles and the degree of loose compaction are determined. The data obtained include the particle size distribution, gradation coefficient, penetration resistance value, and permeability coefficient of soil particles.

[0090] This method uses high-frequency surface radar to obtain the thickness of loose layers, the intensity of reflected waves at the interface, and the propagation velocity of radar waves. Combined with particle size distribution, gradation coefficient, penetration resistance, and permeability coefficient obtained from field sampling and laboratory analysis, it accurately quantifies the seismic physical characteristics of shallow loose structures. It solves the problems of severe seismic wave attenuation and poor wavelet consistency caused by the complex structure of Quaternary shallow soil and rock. It provides reliable data support for the precise delineation of shallow seismic well depth zones and the optimization of source parameters, and significantly improves the accuracy of seismic exploration in complex areas.

[0091] It should be noted that, as Figure 3 As shown, in step two, the construction process of the medium-deep seismic geological condition model includes:

[0092] Step 21, Historical Drilling Data Analysis: Based on existing historical drilling data, quantitatively determine the target stratum depth, coal seam thickness, surrounding rock type and lithological physical properties, including density, porosity and acoustic velocity.

[0093] Step 22: Determine T3 wave characteristics using inter-well logging methods: Use inter-well logging methods such as acoustic logging, density logging, and resistivity logging to determine the acoustic velocity, density, and resistivity of the T3 wave reflection interface. Analyze the intensity of the T3 wave reflected wave, the wave continuity index, and its attenuation ratio of the seismic reflected wave energy of the underlying coal seam to form a shielding effect intensity map, and quantitatively analyze the target formation characteristics and the T3 wave shielding effect.

[0094] Analyzing the reflection characteristics of the T3 wave interface is crucial because the T3 wave interface is typically the interface between the top of the coal seam and the siltstone roof, where there is a significant difference in lithological properties. The high-intensity reflected waves generated at this interface have a significant shielding and energy attenuation effect on the seismic reflected waves from the underlying coal seam. In seismic exploration, this can lead to a reduction in the seismic imaging quality of the target coal seam in the middle and deep layers. Therefore, by quantitatively analyzing the intensity, continuity, and shielding effect of the T3 wave reflection through sonic logging, density logging, and resistivity logging, we can effectively overcome the interference and data quality attenuation problems of the seismic wave field under complex geological conditions, significantly improve the accuracy of target coal seam identification and imaging quality, and provide clear data support for accurately determining the well depth gradient.

[0095] It should be noted that, as Figure 4 As shown, in step three, the method for well depth test zoning based on the data and features obtained in steps one and two includes:

[0096] Step 31: Import the acquired data into the GIS system: Import the data obtained in Step 1, such as the thickness of the loose layer, the intensity of the interface reflected wave, the radar wave propagation speed, the particle distribution, the penetration resistance, and the data obtained in Step 2, such as the burial depth of the medium-deep coal seam, the coal seam thickness, the type of surrounding rock and its lithological physical properties, and the intensity of the T3 wave shielding effect, into the GIS system.

[0097] Step 32, Design test well depth zoning: In the GIS environment, spatial interpolation is used to perform spatial continuity analysis on the imported data to generate a continuous spatial distribution layer. The overlay analysis function of the GIS system is used to perform overlay analysis on the shallow and medium-deep key parameter layers, and regional analysis is performed based on parameter similarity.

[0098] Step 33: Set the well depth gradient for each zone: Combine the data on the thickness of the shallow loose structure and the burial depth of the medium and deep coal seams within the zone to determine the appropriate well depth gradient for each zone, and output a well depth test zone map with clear markings under different surface and geological conditions in the GIS system platform.

[0099] This method integrates data on shallow loose structure thickness, interface reflected wave intensity, radar wave propagation velocity, soil particle distribution, compaction degree, coal seam depth, coal seam thickness, surrounding rock properties, and T3 wave shielding effect intensity into a GIS system. It then uses spatial interpolation and overlay analysis to perform preliminary and fine regional zoning. Based on the differences in the combination of shallow and medium-deep geological conditions, it dynamically determines the well depth gradient, achieving refined distribution and adaptive dynamic feedback adjustment of well depth parameters in different geomorphic regions. This overcomes the difficulty of accurately quantifying the complex surface and geological conditions that make it difficult to adapt well depth parameters, significantly improving the intelligence of well depth zoning and the quality of seismic imaging.

[0100] It should be noted that, in step 33, the specific method for determining each suitable well depth gradient, based on data on various topographic zones, the thickness of shallow loose structures within each zone, and the burial depth of medium-deep coal seams within each zone, includes:

[0101] Bedrock exposed area: The initial well depth gradient is 5m-10m, and the gradient range is dynamically adjusted by ±1m based on the feedback from the real-time updated rock weathering rate model;

[0102] Farmland areas: The initial well depth gradient is set to 10m-15m, and the gradient range is dynamically adjusted by ±1.5m based on feedback from the real-time agricultural cycle change model.

[0103] Weathered material zone: The initial well depth gradient is set to 15m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time weathering process dynamic monitoring model;

[0104] In valley areas: the initial well depth gradient is set to 10m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time geomorphological evolution trend analysis model.

[0105] Due to significant differences in the thickness of shallow loose structures, soil and rock properties, and reflection characteristics of mid-to-deep geological interfaces in bedrock exposed areas, farmland, weathered areas, and valleys, a uniform well depth gradient is difficult to accurately adapt to changes in geological conditions across different regions. Therefore, differentiated well depth gradients are set for different regions based on their geological, geomorphological, and dynamic change characteristics. The well depth range is then dynamically adjusted in real-time based on model feedback to achieve precise matching of well depth parameters to actual geological conditions, ensuring sufficient excitation of seismic reflection wave energy and improving seismic imaging quality and geological information accuracy. This method sets initial well depth gradients for different geomorphological zones and utilizes a real-time feedback mechanism and dynamic fine-tuning method. Based on regional historical geological feature learning models, agricultural cycle change time-series analysis models, weathering process dynamic monitoring models, and geomorphological evolution trend real-time analysis models, the well depth parameters are dynamically adjusted. This solves the problems of traditional manual well depth setting methods failing to respond accurately to changes in surface conditions and exhibiting poor gradient adaptability. It achieves real-time and accurate optimization of well depth parameters, effectively improving seismic data consistency, imaging quality, and geological interpretation reliability, and significantly enhancing seismic exploration efficiency and intelligence.

[0106] To more realistically and meticulously demonstrate the advantages of this method in seismic exploration well depth analysis under complex surface environments, the following four specific examples are presented for detailed comparative testing. Prior to the comparative testing of these examples, the following objective testing conditions were pre-agreed upon to ensure data accuracy and reliability:

[0107] (1) The test area was selected in the area adjacent to the Changping coalfield in the southern part of the Qinshui Basin to ensure that the geological environment conditions, shallow surface structure thickness and coal seam burial depth characteristics were similar in the region.

[0108] (2) All tests used the Geometrics Geode 3D seismic acquisition equipment, with a consistent sampling frequency (1ms) and recording length (2000ms);

[0109] (3) On-site construction personnel receive uniform training and have consistent professional experience to avoid differences in personnel operation;

[0110] (3) All data were collected under sunny weather conditions with no rainfall or significant temperature differences, in order to avoid external interference factors affecting the experimental results.

[0111] Example 1: Bedrock exposed area (area 2.0 km²) 2 )

[0112] Manual group: Using traditional experience, the test was conducted at a fixed well depth of 7.5m. The average data image clarity of 10 test points was 84.5% and the waveform continuity was 79.5%. The energy attenuation of the reflected wave was obvious, and the boundary between the T3 reflected wave and the target coal seam reflected wave was blurred, which affected the imaging quality.

[0113] Test group: In this method, a historical geological feature learning model is used to analyze the regional stability index (0.85) and historical rock weathering rate (average weathering rate of 0.3 cm / a per year) in real time. The RGB brightness is dynamically adjusted to 205 (±8), the color standard deviation is 16 (±3), the texture contrast is 0.72 (±0.08), and the texture entropy is 2.6 (±0.4). The well depth is optimized to 7±0.5m through model feedback. The average imaging clarity of 10 test points is improved to 98.2%, and the waveform continuity is improved to 95.3%, which are 13.7% and 16.5% higher than the manual group, respectively. This method solves the problem of uniform well depth and blurry imaging caused by manual experience and significantly improves data quality.

[0114] Example 2: Farmland area (3.5 km²) 2 )

[0115] The manual group: The depth range of the manual well was randomly set within 10-15m (actually 12m). The average clarity of the seismic imaging at the 10 test points was 78.3%, and the waveform continuity was 74.8%. Due to the significant impact of seasonal changes, the energy loss of seismic waves was obvious.

[0116] Test group: The agricultural cycle variation LSTM time series analysis model was applied, and the color standard deviation was finely adjusted to 27 (±4), regularity coefficient to 0.82 (±0.05), and boundary sharpness index to 0.77 (±0.04) based on crop type (maize) and growth cycle (80 days after sowing). The model optimized the well depth to 11.5±0.75m in real time. The imaging clarity of 10 test points was improved to 93.5%, and the waveform continuity was improved to 91.2%, which were 15.2% and 16.4% higher than those of the manual group, respectively. This effectively solved the problem of unstable seismic wave propagation caused by seasonal disturbances in farmland areas.

[0117] Example 3: Weathered area (area 2.8 km²) 2 )

[0118] Manual group: The traditional method is based on experience and the well depth is fixed at 17m. The average waveform continuity of 10 test points is 69.8% and the signal-to-noise ratio is 63.5%. Due to the large differences in the development degree of weathering fractures, the seismic wave energy attenuates quickly, making it difficult to effectively extract the reflected wave information of the target coal seam.

[0119] Test Group: This method is based on a dynamic monitoring numerical model of weathering process, combined with real-time rainfall (recent cumulative rainfall of 85 mm) and temperature data (annual average temperature of 15℃), adjusting the image color standard deviation to 45 (±5), fracture density to 6 fractures / ㎡ (±0.8 fractures / ㎡), texture entropy to 5.3 (±0.4), and dynamically optimizing the well depth gradient to 17±1 m based on feedback. The waveform continuity of 10 test points was improved to 88.5%, and the signal-to-noise ratio was improved to 89.6%, which are 18.7% and 26.1% higher than the manual group, respectively. This effectively solves the problems of severe energy attenuation and low signal-to-noise ratio in the weathering fracture zone.

[0120] Example 4: Valley area (1.5 km²) 2 )

[0121] Artificial group: The well depth was randomly set within the range of 10-20m (actually 15m). The wave field consistency of the data from 10 test points was 77.8% and the energy resolution was 74.5%. Due to the drastic changes in the valley topography and the complex undulations of the interface, the characteristics of the reflected wave group were not obvious and the energy loss was serious.

[0122] Test group: Using a real-time geomorphological evolution trend analysis model, based on historical erosion (annual erosion average of 0.25 m / a) and recent rainfall data (cumulative rainfall of 60 mm in the last 7 days), the DEM slope was finely adjusted in real time to 16° (±1.5°), depth difference to 2.2 m (±0.3 m), and standard deviation of valley width to 0.4 m (±0.05 m). The well depth gradient was dynamically optimized to 15 ± 0.5 m. The wavefield consistency of the data from 10 test points was improved to 94.3%, and the energy resolution was improved to 92.5%, which are 16.5% and 18% higher than those of the manual group, respectively. This effectively overcomes the problem of discontinuous seismic data and low energy resolution caused by geomorphological changes in valley areas.

[0123] In summary, this invention comprehensively utilizes a dynamic model-driven mechanism driven by multi-source geological data and real-time feedback to accurately adapt to the differentiated characteristics of various landforms, including exposed bedrock areas, farmland, weathered areas, and valleys. It achieves refined and automated dynamic adjustment of well depth gradient parameters, effectively improving the accuracy, reliability, and analytical efficiency of well depth parameter determination in complex surface environments. This effectively overcomes the problems of low efficiency, high fatigue, and insufficient accuracy associated with traditional manual judgment processes, and enhances the quality of seismic wavefield imaging in complex terrain areas. This provides a reliable and high-quality data foundation for subsequent refined seismic sequence stratigraphy, geological inversion analysis, and coalbed methane exploration prediction.

[0124] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0125] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for seismic exploration well depth analysis based on complex surface environments, characterized in that, Includes the following steps: Step 1: Analyze complex surface and shallow geological conditions. This involves acquiring orthophotos and digital elevation models of the surface through aerial photography to identify landform features. High-frequency surface radar is used to detect the thickness and distribution of loose structures in the shallow layer. Field sampling and laboratory analysis clarify the distribution of soil and rock particles and the degree of compaction. The specific steps for image processing, including acquiring high-precision orthophotos and digital elevation models of the surface using drones, identifying landform features, and achieving preliminary area segmentation, are as follows: Step 11: Perform radiometric and orthorectification corrections on the acquired aerial image data; Step 12: Using the Mean-shift algorithm, based on color similarity and spatial proximity, the original image is segmented into several homogeneous region patches; Step 13: Extract the image quantization features of each segmented region and initially establish an image feature library for different landform categories; Step 14: Classify the initially segmented regions using the random forest classification algorithm, and preliminarily identify the types of various landform regions based on the classification results; Step 15: Adaptive image refinement processing is performed on the initially identified areas: A machine learning model is constructed based on historical geological stability index and rock weathering rate. The RGB brightness, color standard deviation, texture contrast, and texture entropy threshold are dynamically adjusted. The threshold is optimized through feedback results of regional features. The segmentation boundary of the exposed bedrock area is finely adjusted. A time series analysis model of seasonal agricultural cycle changes is constructed. The color standard deviation, regularity coefficient, and boundary sharpness index are dynamically adjusted. Farmland boundaries and crop types are finely segmented and identified. Based on regional historical weathering rate and real-time climate monitoring data, a dynamic data value analysis model is constructed. The fissure identification density and texture entropy index are dynamically adjusted. The contour boundary of the weathered area is finely adjusted. Combined with digital elevation data, a geomorphic evolution trend analysis model is constructed. The slope, geomorphic depth difference, and valley width threshold are adjusted in real time to finely depict the contour details of the valley area. Step 2: Construct a medium-deep geological model: Determine the burial depth, thickness, and surrounding lithology of the target strata by integrating historical borehole data; use inter-well logging data to determine the characteristics of the T3 wave reflection interface and its shielding effect on the coal seam; and establish a medium-deep seismic geological model. Step 3, Designing Zoned Well Depth Tests: Based on the analysis results of Step 1 and Step 2, the region is divided using a geographic information system, and several different levels of well depth gradients are designed for different seismic wavefield characteristics. Step 4: Collect multi-gradient well depth test data: Conduct seismic excitation tests according to well depth gradients in each test zone, and collect shallow and medium-deep seismic reflection data using nodal three-dimensional seismic sensors; Step 5, Automatic Seismic Reflection Data Analysis Based on Deep Learning: Construct an automatic identification system based on convolutional neural networks, train the model using labeled training data, automatically extract waveform continuity, signal-to-noise ratio, and attenuation rate using the trained automatic identification system, score the data, and optimize the model using transfer learning. Step 6, Well Depth Intelligent Optimization Inversion: Establish a machine learning mapping model with the model evaluation results from Step 5 and the corresponding well depth data, and use the XGBoost intelligent optimization algorithm to perform well depth inversion analysis to determine the optimal well depth parameters for each partition.

2. The method for seismic exploration well depth analysis based on complex surface environments according to claim 1, characterized in that, The method further includes: Step 7, Field verification of optimized well depth and construction of 3D data volume: Use the optimized well depth data to conduct actual field tests, re-acquire data, use data processing software to establish a 3D seismic time profile, analyze wavefield characteristics and wavelet consistency, use 3D seismic data to refine wavefield imaging, and clarify the characteristics of shallow and medium-deep layers. Step 8, Geological Application Analysis: Based on the optimized 3D data volume, conduct detailed seismic sequence grid construction and geological inversion analysis to clarify the precise distribution and occurrence conditions of coal seams, analyze geological structural changes and coalbed methane occurrence characteristics, and provide a geological structural interpretation report.

3. The method for seismic exploration well depth analysis based on complex surface environments according to claim 2, characterized in that, In step one, orthophotos and digital elevation models of the land surface are acquired through aerial photography. The specific method for identifying landform features is to use supervised classification to obtain input feature data from the images and identify various landform types, including: Image regions with RGB brightness greater than 200, color standard deviation less than 20, texture contrast greater than 0.7 and texture entropy less than 3 are identified as bedrock exposed areas. Image regions with an RGB color standard deviation of less than 30, a regularity coefficient of greater than 0.8, and a boundary sharpness index of greater than 0.75 are identified as farmland areas. Image regions with an RGB color standard deviation greater than 40, a crack density greater than 5 lines / m², and a texture entropy greater than 5 are identified as weathered areas. Image regions with a DEM slope greater than 15°, a terrain depth difference greater than 2 meters, and a valley width standard deviation less than 0.5m are identified as valley areas.

4. The method for seismic exploration well depth analysis based on complex surface environments according to claim 3, characterized in that, In step one: The input feature data of the bedrock exposed area is dynamically adjusted by a regional historical geological feature learning model, with RGB brightness ±10, color standard deviation ±5, texture contrast ±0.1, and texture entropy ±0.

5. The input feature data of farmland plots are adaptively adjusted using an agricultural cycle variation model, with color standard deviation ±5, regularity coefficient ±0.1, and boundary sharpness index ±0.

05. The input feature data of the weathered area are dynamically adjusted by the weathering process dynamic monitoring model, with color standard deviation ±5, crack density ±1 line / m², and texture entropy ±0.

5. In the valley area, the slope ±2°, the depth difference ±0.5m, and the standard deviation of the valley width ±0.1m are dynamically adjusted using a geomorphic evolution trend analysis model.

5. The method for seismic exploration well depth analysis based on complex surface environments according to claim 4, characterized in that, The learning model for the historical geological features of the region is a machine learning model constructed using the regional stability index and the historical rock weathering rate. The agricultural cycle variation model is a time series analysis model established based on seasonal crop planting patterns, crop types, and growth cycles. The dynamic monitoring model for the weathering process is a numerical model based on historical data analysis, which is based on real-time rainfall, temperature, and historical weathering rate. The geomorphic evolution trend analysis model is a real-time analysis model built based on historical erosion, rainfall data, and geomorphic evolution trends.

6. The method for seismic exploration well depth analysis based on complex surface environments according to claim 5, characterized in that, In step one, high-frequency surface radar is used to detect the thickness and spatial distribution characteristics of the shallow loose structure. The acquired characteristic data includes the thickness of the loose layer, the intensity value of the reflected wave at the interface of the loose layer, and the distribution of radar wave propagation velocity. Through field sampling and laboratory analysis, the distribution of soil particles and the degree of loose compaction are determined. The acquired data includes the particle size distribution, gradation coefficient, penetration resistance value, and permeability coefficient of the soil particles.

7. The method for seismic exploration well depth analysis based on complex surface environments according to claim 1, characterized in that, In step two, the construction process of the intermediate-deep geological model includes: Step 21, Historical Drilling Data Analysis: Based on existing historical drilling data, quantitatively determine the target stratum depth, coal seam thickness, surrounding rock type and lithological physical properties, including density, porosity and acoustic velocity. Step 22: Determine T3 wave characteristics using inter-well logging methods: Use inter-well logging methods such as acoustic logging, density logging, and resistivity logging to determine the acoustic velocity, density, and resistivity of the T3 wave reflection interface. Analyze the intensity of the T3 wave reflected wave, the wave continuity index, and its attenuation ratio of the seismic reflected wave energy of the underlying coal seam to form a shielding effect intensity map, and quantitatively analyze the target formation characteristics and the T3 wave shielding effect.

8. The method for seismic exploration well depth analysis based on complex surface environments according to claim 7, characterized in that, In step three, well depth testing and zoning based on the data and features obtained in steps one and two includes: Step 31: Import the acquired data into the GIS system: Import the data obtained in Step 1, such as the thickness of the loose layer, the intensity of the interface reflected wave, the radar wave propagation speed, the particle distribution, the penetration resistance, and the data obtained in Step 2, such as the burial depth of the medium-deep coal seam, the coal seam thickness, the type of surrounding rock and its lithological physical properties, and the intensity of the T3 wave shielding effect, into the GIS system. Step 32, Design test well depth zoning: In the GIS environment, spatial interpolation is used to perform spatial continuity analysis on the imported data to generate a continuous spatial distribution layer. The overlay analysis function of the GIS system is used to perform overlay analysis on the shallow and medium-deep key parameter layers, and regional analysis is performed based on parameter similarity. Step 33: Set the well depth gradient for each zone: Combine the data on the thickness of the shallow loose structure and the burial depth of the medium and deep coal seams within the zone to determine the appropriate well depth gradient for each zone, and output a well depth test zone map with clear markings under different surface and geological conditions in the GIS system platform.

9. The method for seismic exploration well depth analysis based on complex surface environments according to claim 8, characterized in that, In step 33, by combining data on various topographic zones, the thickness of shallow loose structures within each zone, and the burial depth of medium-deep coal seams within each zone, the appropriate well depth gradient is determined. Specific methods include: Bedrock exposed area: The initial well depth gradient is 5m-10m, and the gradient range is dynamically adjusted by ±1m based on the feedback from the real-time updated rock weathering rate model; Farmland areas: The initial well depth gradient is set to 10m-15m, and the gradient range is dynamically adjusted by ±1.5m based on feedback from the real-time agricultural cycle change model. Weathered material zone: The initial well depth gradient is set to 15m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time weathering process dynamic monitoring model; In valley areas: the initial well depth gradient is set to 10m-20m, and the gradient range is dynamically adjusted by ±2m based on the feedback from the real-time geomorphological evolution trend analysis model.

Images