Artificial intelligence-based underground fault visualization system and method

Abstract

An artificial intelligence-based underground fault visualization method includes steps as follows. The geological and geophysical underground fault data are integrated through the generative artificial intelligence to establish a basic fault model; the measurement locations of the electrical resistivity tomography are marked, and the measurement locations of the electrical resistivity tomography are combined with the basic fault model; the data of the measurement locations of the electrical resistivity tomography are inverted to generate multiple two-dimensional cross-sections of the underground faults; the multiple two-dimensional cross-sections are preprocessed; and the multiple two-dimensional cross-sections of the underground faults are integrated into a three-dimensional model through the generative artificial intelligence, in which the step of integrating the multiple two-dimensional cross-sections includes performing spatial interpolation between the multiple two-dimensional cross-sections, and the three-dimensional model provides the visualization of the underground faults.

Description

[Technical Field]

[0001] This invention relates to a computer system and a method for visualizing underground faults, and more particularly to an artificial intelligence-based system and method for visualizing underground faults. [Previous Technology]

[0002] In recent years, earthquakes have occurred frequently, posing a significant threat to people's lives and property. The occurrence of earthquakes is closely related to the activity of underground faults. Therefore, monitoring changes in fault structure is crucial for earthquake early warning and disaster prevention.

[0003] However, traditional geological survey and fault analysis methods are time-consuming and labor-intensive, requiring a large investment of human and financial resources, making it difficult to meet the needs for immediacy and accuracy. [Summary of the Invention]

[0004] This invention proposes an artificial intelligence-based system and method for visualizing underground faults, which improves upon the problems of previous technologies.

[0005] In some embodiments of the present invention, the artificial intelligence-based underground fault visualization system proposed in this invention includes a storage device and a processor. The processor is electrically connected to the storage device, and the storage device stores at least one instruction. The processor is used to access and execute at least one instruction to: integrate geological and geophysical underground fault information through generative artificial intelligence to establish a basic fault model; mark the measurement locations of resistivity image detection and combine the data of the measurement locations of the resistivity image detection with the basic fault model; invert the data of the measurement locations of the resistivity image detection to generate multiple two-dimensional cross-sectional images of the underground fault; and integrate the multiple two-dimensional cross-sectional images of the underground fault into a three-dimensional model through generative artificial intelligence, the three-dimensional model providing visualization of the underground fault.

[0006] In some embodiments of the present invention, the processor is configured to access and execute at least one instruction to: identify fault structure features by comparing the measurement location data of ground resistance image detection at different locations through generative artificial intelligence, and interpret and analyze the three-dimensional fault model.

[0007] In some embodiments of the present invention, the processor is configured to access and execute at least one instruction to: combine data of the measurement locations detected by ground resistance images at different times, analyze the changes and movement directions of the underground fault over time, so as to infer the possible location of future earthquakes.

[0008] In some embodiments of the present invention, the processor is configured to access and execute at least one instruction to: learn from hydrogeological data, historical disaster records, and public transportation data through generative artificial intelligence to establish a risk prediction model, the risk prediction model being used to monitor changes in hydrogeological data in real time to assess potential risks in advance, wherein the hydrogeological data includes soil liquefaction potential, seismic data, and data on the measurement locations of ground resistivity image detection.

[0009] In some embodiments of the present invention, the risk prediction model provides customized risk assessments based on the characteristics of different public transport routes and stations.

[0010] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults proposed in the present invention includes the following steps: integrating geological and geophysical underground fault information through generative artificial intelligence to establish a basic fault model; marking the measurement locations of resistivity image detection and combining the data of the measurement locations of the resistivity image detection with the basic fault model; inverting the data of the measurement locations of the resistivity image detection to generate multiple two-dimensional cross-sectional images of the underground fault; preprocessing the multiple two-dimensional cross-sectional images by at least one of the following: (a) registering the multiple two-dimensional cross-sectional images based on the measurement locations and orientations; (b) applying noise reduction; (c) applying contrast enhancement; and integrating the multiple two-dimensional cross-sectional images of the underground fault into a three-dimensional model through generative artificial intelligence, wherein the step of integrating the multiple two-dimensional cross-sectional images includes performing spatial interpolation between the multiple two-dimensional cross-sectional images, and the three-dimensional model provides visualization of the underground fault.

[0011] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults further includes: using generative artificial intelligence to compare the measurement data of the ground resistance image detection at different locations, identify fault structure features, and interpret and analyze the three-dimensional fault model.

[0012] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults further includes: combining data from the measurement locations detected by ground resistance images at different times, analyzing the changes and movement directions of underground faults over time, in order to predict the possible locations of future earthquakes.

[0013] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults further includes: learning from hydrogeological data, historical disaster records, and public transportation data through generative artificial intelligence to establish a risk prediction model. The risk prediction model is used to monitor changes in hydrogeological data in real time to assess potential risks in advance. The hydrogeological data includes soil liquefaction potential, seismic data, and data on the measurement locations of ground resistance image detection.

[0014] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults further includes: providing customized risk assessments based on the characteristics of different public transport routes and stations through a risk prediction model.

[0015] In summary, the technical solution of the present invention has significant advantages and beneficial effects compared with the prior art. By utilizing the AI-based underground fault visualization system and method of the present invention, generative artificial intelligence combined with Earth Resistance Image Detection (ERT) technology is employed to accelerate and optimize underground fault investigation and analysis. By integrating the latest underground fault information, establishing a basic fault model, and visualizing the measurement locations of ERT measurements, the present invention provides more comprehensive and accurate visualized information on underground faults, which is helpful for earthquake early warning, disaster prevention policy formulation, infrastructure safety assessment, and environmental disaster prevention.

[0016] The above description will be described in detail below with reference to embodiments, and the technical solution of the present invention will be further explained.

Implementation Method

[0017] To make the description of the present invention more detailed and complete, reference can be made to the accompanying drawings and the various embodiments described below, in which the same numbers represent the same or similar elements. On the other hand, well-known elements and steps are not described in the embodiments to avoid unnecessarily limiting the present invention.

[0018] Referring to Figure 1, the technical embodiment of the present invention is an artificial intelligence-based underground fault visualization system 100, which can be applied to computers or widely used in related technical aspects. This artificial intelligence-based underground fault visualization system 100 achieves considerable technological advancement and has broad industrial application value. The specific implementation of the artificial intelligence-based underground fault visualization system 100 will be described below with reference to Figure 1.

[0019] It should be understood that various embodiments of the AI-based underground fault visualization system 100 are described in conjunction with Figure 1. In the following description, for ease of explanation, numerous specific details are further provided to offer a comprehensive description of one or more embodiments. However, the technology can be implemented without these specific details. In other examples, known structures and devices are shown in block diagram form for the effective description of these embodiments. The term "by way of example" as used herein means "as an example, instance, or illustration." Any embodiment described herein as "by way of example" is not to be construed as preferred or superior to other embodiments.

[0020] In practice, in some embodiments of the present invention, the AI-based underground fault visualization system 100 can be a server, a computer host, or other computer device. Regarding servers, many technologies have been developed or are under development to manage the operation of computer servers, generally providing accessibility, consistency, and efficiency. Remote management allows for the removal of input / output interfaces for servers (e.g., display screens, mice, keyboards, etc.) and the need for network administrator entities to access each server. For example, large data centers containing many computer servers typically use various remote management tools to manage, configure, monitor, and troubleshoot server hardware and software.

[0021] It should be understood that the terms "about," "approximately," or "roughly" used herein are used to modify any quantity that may vary slightly, but such slight variations do not change its nature. Unless otherwise specified in the embodiments, the error range of the values ​​modified by "about," "approximately," or "roughly" is generally permissible within 20 percent, preferably within 10 percent, and more preferably within 5 percent.

[0022] In practice, in some embodiments of the present invention, the AI-based underground fault visualization system 100 may selectively establish a connection with the user terminal device 190. It should be understood that, in the embodiments and claims, the description of "connection" can broadly refer to a component indirectly communicating with another component via wired and / or wireless means, or a component not physically connected to another component without needing to be connected via other components. For example, the user terminal device 190 may be a mobile phone, computer, or similar device. Users can connect to the AI-based underground fault visualization system 100 through an application (APP) or web browser on the user terminal device 190 to obtain visualization information about underground faults.

[0023] Figure 1 is a block diagram of an artificial intelligence-based underground fault visualization system 100 according to an embodiment of the present invention. As shown in Figure 1, the artificial intelligence-based underground fault visualization system 100 includes a storage device 110, a processor 120, and a communication device 130. For example, the storage device 110 may be a hard drive, a flash storage device, or other storage medium, the processor 120 may be a central processing unit, and the communication device 130 may be a wired and / or wireless network device.

[0024] In terms of architecture, the storage device 110 is electrically connected to the processor 120, and the processor 120 is electrically connected to the communication device 130. The communication device 130 can perform wired and / or wireless communication with the user terminal device 190. It should be understood that, in the embodiments and claims, the description involving "electrical connection" can generally refer to one component being indirectly electrically coupled to another component through other components, or one component being directly electrically connected to another component without needing to go through other components. For example, the storage device 110 can be a built-in storage device directly electrically connected to the processor 120, or the storage device 110 can be an external storage device indirectly connected to the processor 120 through a network device.

[0025] In some embodiments of the present invention, the storage device 110 stores at least one instruction. The processor 120 is used to access and execute at least one instruction to: integrate geological and geophysical subsurface fault information through generative artificial intelligence to establish a basic fault model; mark the measurement locations of resistivity image detection and combine the data of the measurement locations of the resistivity image detection with the basic fault model; invert the data of the measurement locations of the resistivity image detection to generate multiple two-dimensional cross-sectional images of the subsurface fault; and integrate the multiple two-dimensional cross-sectional images of the subsurface fault into a three-dimensional model through generative artificial intelligence, the three-dimensional model providing visualization of the subsurface fault. Thus, the AI-based subsurface fault visualization system 100 of the present invention provides more comprehensive and accurate visualization information of subsurface faults, which is helpful for earthquake early warning, disaster prevention policy formulation, infrastructure safety assessment, and environmental disaster prevention.

[0026] In some embodiments of the present invention, the processor 120 is configured to access and execute at least one instruction to: identify fault structure features by comparing the measurement location data of ground resistance image detection at different locations through generative artificial intelligence, and interpret and analyze the three-dimensional fault model.

[0027] In some embodiments of the present invention, the processor 120 is configured to access and execute at least one instruction to: combine data of the measurement locations detected by ground resistance images at different times, analyze the changes and movement directions of the underground fault over time, so as to infer the possible location of future earthquakes.

[0028] In some embodiments of the present invention, the processor 120 is configured to access and execute at least one instruction to: learn from hydrogeological data, historical disaster records, and public transportation data through generative artificial intelligence to establish a risk prediction model, the risk prediction model being used to monitor changes in hydrogeological data in real time to assess potential risks in advance, wherein the hydrogeological data includes soil liquefaction potential, seismic data, and data on the measurement locations of ground resistivity image detection.

[0029] In some embodiments of the present invention, the risk prediction model provides customized risk assessments based on the characteristics of different public transport routes and stations.

[0030] To further illustrate the operation of the artificial intelligence-based underground fault visualization system 100 described above, please refer to Figures 1 and 2. Figure 2 is a flowchart of an artificial intelligence-based underground fault visualization method 200 of the artificial intelligence-based underground fault visualization system 100 according to an embodiment of the present invention. As shown in Figure 2, the artificial intelligence-based underground fault visualization method 200 includes steps S201 to S204 (it should be understood that, unless otherwise specified, the order of the steps mentioned in this embodiment can be adjusted according to actual needs, and they can even be executed simultaneously or partially simultaneously).

[0031] The AI-based method for visualizing underground faults 200 can take the form of a computer program product on a non-transitory computer-readable recording medium, which has a plurality of computer-readable instructions contained in the medium. Suitable recording media may include any of the following: non-volatile memory, such as read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electronically eraseable programmable read-only memory (EEPROM); volatile memory, such as static random access memory (SRAM), dynamic random access memory (DRAM), double data rate random access memory (DDR-RAM); optical storage devices, such as read-only optical discs (CD-ROM), read-only bit multi-function audio-visual discs (DVD-ROM); magnetic storage devices, such as hard disk drives, floppy disk drives.

[0032] In step S201, geological and geophysical information on underground faults is integrated using generative artificial intelligence to establish a basic fault model. In step S202, the measurement locations of the resistivity image detection are marked, and the data from the measurement locations of the resistivity image detection are combined with the basic fault model. In step S203, the data from the measurement locations of the resistivity image detection are inverted to generate multiple two-dimensional cross-sectional images of the underground fault. The multiple two-dimensional cross-sectional images are preprocessed by at least one of the following: (a) registering multiple two-dimensional cross-sectional images based on measurement locations and orientations; (b) applying noise reduction; (c) applying contrast enhancement. In step S204, the multiple two-dimensional cross-sectional images of the underground fault are integrated into a three-dimensional model using generative artificial intelligence. The step of integrating multiple two-dimensional cross-sectional images includes performing spatial interpolation between the multiple two-dimensional cross-sectional images. The three-dimensional model (e.g., an interactive three-dimensional underground fault model) provides visualization of the underground fault. Therefore, the AI-based underground fault visualization method 200 of the present invention provides more comprehensive and accurate visualization information of underground faults, which is helpful for earthquake early warning, disaster prevention policy formulation, infrastructure safety assessment, and environmental disaster prevention.

[0033] In practice, for example, the AI-based subsurface fault visualization method 200 combines generative AI (using Gemini Advanced as an example, but not limited to it) with geoelectric resistivity image detection technology, achieving a comprehensive upgrade in subsurface fault investigation and analysis. Step S201: Data integration and model building: Utilizing generative AI, the latest subsurface fault information from different sources is rapidly integrated, including geological survey reports, seismic observation data, geological models, etc. Based on this, a basic fault model is established, covering information such as fault type, depth, extent, and angle. Step S202: Combining ground feature marking with geoelectric resistivity image detection data: Using an appropriate programming environment, the measurement locations of geoelectric resistivity image detection are marked, and the data from these measurement locations are combined with the basic fault model. Visualization and analysis in steps S203 and S204: Using geophysical inversion software tools (such as PyGimli software), the measurement location data of the ground resistivity image detection is inverted to generate a two-dimensional cross-sectional map of the underground fault. Then, using the powerful computing capabilities of generative artificial intelligence, multiple two-dimensional cross-sectional maps are integrated into a three-dimensional model to realize the visualization of the underground fault.

[0034] In some embodiments of the present invention, the artificial intelligence-based underground fault visualization method 200 further includes: using generative artificial intelligence to compare the measurement data of ground resistance image detection at different locations, identifying fault structure features, and interpreting and analyzing the three-dimensional fault model. In practice, for example, the fault structure interpretation and analysis of the artificial intelligence-based underground fault visualization method 200 involves: using the analysis function of the generative artificial intelligence to compare the data maps of ground resistance image detection at different locations, identifying fault structure features (such as fault dip angle, strike, curvature, etc.), and interpreting and analyzing the three-dimensional fault model.

[0035] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults 200 further includes: combining data from the measurement locations detected by ground resistance images at different times, analyzing the changes and movement directions of the underground faults over time, in order to predict the possible locations of future earthquakes.

[0036] In summary, the AI-based visualization method for underground faults 200 is innovative, as illustrated below: Cross-domain integration: Combining generative AI technology with geophysical exploration technology, it achieves cross-domain integration of underground fault investigation and analysis. High efficiency and automation: Utilizing the powerful computing and analytical capabilities of generative AI, it automates the processing of large amounts of data, significantly improving work efficiency. Visual presentation: It presents complex underground fault structures as three-dimensional models, making them intuitive and easy to understand, which is helpful for decision-making. Predicting potential risks: Through time-series analysis, it predicts the possible locations of future earthquakes, providing a scientific basis for earthquake early warning.

[0037] Examples of achievements regarding the AI-based visualization method for underground faults are as follows: Interactive 3D Underground Fault Model: Combining data from ground resistivity imagery with fault information, the model presents the detailed structure of underground faults in three dimensions. Users can observe the fault's strike, dip angle, and depth from different perspectives and interact with the model, such as zooming in, zooming out, and rotating. Earthquake Simulation and Risk Assessment: Combining historical earthquake data with fault models, the model simulates earthquake occurrence and surface vibration under different scenarios to assess the earthquake risk level of specific areas. Underground Fault Characteristic Analysis Tool: A generative AI-based analysis tool has been developed to automatically extract fault features (such as fault length, width, and offset) and provide analysis results on fault activity and seismic potential. Popular Science Education Platform: An interactive popular science education platform has been established to display 3D fault models and earthquake simulation results, and to provide relevant popular science knowledge to improve public understanding of earthquakes and disaster prevention awareness.

[0038] In practice, for example, the AI-based subsurface fault visualization method 200 can employ various technologies to achieve the above results. A suitable programming environment or cloud platform: as the primary development environment, it utilizes the provided graphics processing unit (GPU) resources to accelerate model training and data processing. Python, combined with relevant map visualization libraries (such as Folium), can be used to overlay the measurement locations of resistivity image detection with fault models on a map, facilitating users' understanding of the geographical distribution of faults. Geophysical inversion software tools (such as PyGIMLi) can be used to process the data from resistivity image detection, perform resistivity inversion, and construct subsurface fault models. Generative AI (such as Google Gemini Advanced): utilizing its powerful natural language processing capabilities, it enables human-computer interaction, allowing users to ask questions in natural language to obtain fault information, earthquake risk assessments, etc.; utilizing its image generation capabilities, it transforms data into intuitive 3D models and earthquake simulation animations; utilizing its analytical capabilities, it extracts fault features from the data of resistivity image detection for earthquake risk assessment.

[0039] Specifically, the artificial intelligence-based method for visualizing subsurface faults 200 can integrate fault models with map information using Geographic Information System (GIS) related technologies (such as Folium); process data from geoelectric resistivity image detection and construct models using geophysical inversion software tools (such as PyGIMLi); and realize model visualization, interactive functions, earthquake simulation, risk assessment, science education platform, etc. using generative artificial intelligence.

[0040] In practice, the AI-based underground fault visualization method 200 has a wide range of applications, as exemplified below: Geological survey and monitoring: Rapidly establish accurate fault models, monitor fault activity, and predict potential earthquake risks. Earthquake early warning and disaster prevention: Provide real-time earthquake early warnings based on changes in fault activity to assist governments and the public in taking response measures. Infrastructure safety assessment: Assess the safety risks of critical infrastructure (such as bridges, roads, and buildings) in potential earthquakes and propose improvement recommendations.

[0041] In practice, the AI-based underground fault visualization method 200 has multiple benefits, as exemplified below: Improved efficiency and accuracy: Accelerates underground fault investigation and analysis, providing more accurate fault information. Reduced disaster losses: Reduces loss of life and property caused by earthquakes through earthquake early warning and disaster prevention measures; Enhances infrastructure safety; Ensures critical infrastructure remains functional during earthquakes, protecting public safety. Promotes environmental disaster prevention: Predicts and prevents environmental disasters caused by earthquakes, protecting the natural environment.

[0042] In practice, the AI-based underground fault visualization method 200 can meet the needs of different users, for example as follows: Government: Provides regional risk assessments to assist in the formulation of disaster prevention policies. Experts: Provides detailed geological change data to assist in research and analysis. Residents: Provides simplified housing safety assessment reports to help determine whether repairs or relocation are necessary. Insurance companies: Provides loss assessment reports.

[0043] In some embodiments of the present invention, the artificial intelligence-based method for visualizing underground faults 200 further includes: learning from hydrogeological data, historical disaster records, and public transportation data through generative artificial intelligence to establish a risk prediction model. The risk prediction model is used to monitor changes in hydrogeological data in real time to assess potential risks in advance. The hydrogeological data includes soil liquefaction potential, seismic data, and data on the measurement locations of ground resistance image detection.

[0044] In some embodiments of the present invention, the artificial intelligence-based underground fault visualization method 200 further includes: providing customized risk assessment based on the characteristics of different public transport routes and stations through a risk prediction model.

[0045] In practice, for example, the AI-based underground fault visualization method 200 converts data into HTML code and then presents it as a webpage through an online code hosting platform (such as GitHub). Data can be expressed in different ways and submitted to generative AI (such as Google Gemini, ChatGPT), with formats such as HTML, XML, JSON, and even Excel. In some embodiments, HTML is used to express multiple data items, including location, date and time, text, numbers, and images. In some embodiments, when the data volume is large, XML or JSON data representation can be used. The AI-based underground fault visualization method 200 can process the data from geophysical inversion software tools for geoelectric resistivity image detection.

[0046] Then, an example of a risk prediction model is as follows. The user device 190 submits the following question: If you are an insurance real estate company, how do you assess and quantify the earthquake risk in the northern, southern, western, eastern, and central areas of Hualien to estimate future real estate premiums? How can you combine this with ERT (Earth Resistance Image Detection) measurements of resistivity obtained from the Ruisui Fault (RIP01-RIP10) and the Heyuli Fault (YL01-YL19) to assist insurance real estate companies in assessing the risk of ground damage to real estate caused by soil liquefaction? The answer from the AI-based underground fault visualization system 100: The assessment of earthquake risk in the Hualien area is closely related to soil liquefaction risk. The ERT method can measure the distribution of underground resistivity, which helps to determine the potential for soil liquefaction. The following explains how this information can be combined to assist insurance real estate companies in assessing earthquake risk in various areas of Hualien and estimating future real estate premiums.

[0047] In practice, for example, the risk prediction model can be an earthquake risk assessment model. The AI-based underground fault visualization method 200 can establish an earthquake risk index model based on historical earthquake data (such as the magnitude, intensity, depth, and epicentral distance of historical earthquakes in each region), fault data (such as the distribution, activity, and distance of faults in each region), and geological conditions (such as soil type and groundwater level) to provide an earthquake risk index and quantify the earthquake risk in each region. In addition, the AI-based underground fault visualization method 200 can establish a soil liquefaction risk index model based on resistivity data obtained from ground resistance image detection (such as analyzing ground resistance image detection data of faults to assess soil liquefaction potential), soil characteristics (such as soil particle size, density, and water content), and groundwater level data to provide a soil liquefaction risk index and quantify the soil liquefaction risk in each region. Then, based on the AI-powered underground fault visualization method 200, combined with the earthquake risk index and the soil liquefaction risk index, a seismic risk zoning map of each region was drawn, indicating the risk level of each region.

[0048] In practice, for example, a risk prediction model can be a real estate premium estimation model. The real estate premium estimation model obtains building data (such as building type, structure, floors, age, value, etc.). Based on an earthquake risk zoning map, the model sets risk adjustment coefficients for each zone, and further sets risk adjustment coefficients for different buildings based on the building data. The model sets a base premium based on market conditions or past claims data, and then multiplies the base premium by the earthquake risk adjustment coefficient and then by the building risk adjustment coefficient to obtain the real estate premium.

[0049] In practice, for example, the AI-based underground fault visualization method 200 can perform data analysis and application of ground resistivity image detection, as shown in the following examples: Data processing: Inverting the ground resistivity image detection data of faults in various areas (such as the Ruisui Fault and the Yuli Fault) to obtain the underground resistivity distribution. Soil liquefaction potential assessment: Low resistivity areas may represent high water content and loose sediments, with a higher risk of soil liquefaction; high resistivity areas may represent low water content and dense sediments, with a lower risk of soil liquefaction. Risk map correction: Using ground resistivity image detection data to correct the soil liquefaction risk index model, improving the accuracy of the assessment. Specific location risk assessment: For specific land areas, more accurate soil liquefaction risk assessment can be conducted using measurement point data from nearby ground resistivity image detection.

[0050] To further illustrate how generative artificial intelligence (AI) can be used to process multiple two-dimensional cross-sectional images at different locations (e.g., two-dimensional (2D) geoelectric resistivity (ERT) impedance images) to accelerate the computational evaluation of three-dimensional (3D) fault features, please refer to Figures 1 to 3. Figure 3 is a schematic diagram of a three-dimensional model 300 (i.e., a fault three-dimensional model) according to one embodiment of the present invention.

[0051] The generative artificial intelligence (AI) of the artificial intelligence-based subsurface fault visualization system 100 has three-dimensional (3D) interpretation capabilities. This invention utilizes generative artificial intelligence (AI) to not only integrate subsurface fault information and geoelectric resistivity image detection data, but also to infer the three-dimensional (3D) geometry and characteristics of the fault from multiple two-dimensional (2D) cross-sectional images 301, such as strike 320°, dip angle 310°, and depth.

[0052] Regarding the reasoning process of the generative artificial intelligence (AI) of the artificial intelligence-based underground fault visualization system 100, the generative artificial intelligence can analyze resistivity images, identify fault planes, calculate dip angles and strikes, and integrate information from different profiles. The following will be described in detail with examples.

[0053] The AI-based underground fault visualization system 100 has data integration capabilities. In addition to ground resistance image detection data, the AI-based underground fault visualization system 100 of the present invention can integrate other geological and geophysical data, such as drilling data, seismic data, geological maps, etc., to improve the accuracy of the three-dimensional (3D) fault model.

[0054] Regarding how to calculate three-dimensional (3D) fault features, such as dip angle 310°, strike angle 320°, and depth, from multiple two-dimensional (2D) resistivity images, for example, dip angle 310°: the angle between the fault plane and the horizontal plane, indicating the degree of fault inclination, such as 0 to 90 degrees. Strike angle 320°: the direction of the intersection of the fault plane and the horizontal plane, indicating the direction of fault extension; its strike angle 321 can be 0 to 360 degrees. Depth: the depth of the fault underground, usually referring to the depth of the top or bottom of the fault.

[0055] Since two-dimensional (2D) resistivity images only provide information about the fault on a single cross-section, it is necessary to combine cross-sections from multiple directions to infer the geometry of a three-dimensional (3D) fault. This requires overcoming the following challenges: Data registration: Ensuring the correct spatial relative positions of images from different cross-sections. Data sparsity: There may be gaps between different cross-sections, requiring interpolation or inference. Data uncertainty: Two-dimensional (2D) images themselves have issues such as resolution and noise, which can affect the accuracy of parameter calculations.

[0056] In order to overcome the above challenges, the method of generating a three-dimensional (3D) fault model from multiple two-dimensional (2D) resistivity images using generative artificial intelligence (AI) in the subsurface fault visualization system 100 based on artificial intelligence includes data preprocessing, generative artificial intelligence fault identification and parameter extraction, multi-profile integration, depth estimation and strike estimation, which are detailed below.

[0057] 1. Data Preprocessing: Each two-dimensional (2D) resistivity image is registered according to its measurement location and orientation using Global Positioning System (GPS) information or ground control points. The resistivity images are then processed for noise reduction and enhancement, such as using filtering and contrast enhancement methods.

[0058] 2. Generative Artificial Intelligence (AI) Fault Recognition and Parameter Extraction: Geometric parameters of faults are extracted using pre-designed prompts. The AI ​​model can learn fault features in resistivity images, such as abrupt changes in resistivity and linear structures, to automatically identify faults and calculate the dip angle 31° and depth of faults on each profile.

[0059] 3. Multi-section integration: Fault parameters from different sections are integrated into a three-dimensional (3D) space using generative artificial intelligence (AI), for example, by using spatial interpolation or surface reconstruction techniques to generate a continuous 3D fault model. During the integration process, generative artificial intelligence (AI) can be used to analyze the relationships between different sections, such as the changing trend of the fault strike, thereby more accurately estimating the 3D geometry of the fault.

[0060] 4. Depth estimation: Based on the fault depth information on multiple profiles, the depth of the fault at different locations is estimated using methods such as spatial interpolation or fitting.

[0061] 5. Strike estimation: Based on the location and dip direction of the fault on multiple cross-sections, the strike 320 of the fault is inferred. For example, vector analysis, trend surface analysis, and other methods can be used to calculate the strike 320 of the fault.

[0062] In summary, the technical solution of the present invention has significant advantages and beneficial effects compared with the prior art. By utilizing the AI-based underground fault visualization system 100 and AI-based underground fault visualization method 200 of the present invention, generative artificial intelligence is combined with ground resistance image detection technology to accelerate and optimize underground fault investigation and analysis. By integrating the latest underground fault information, establishing a basic fault model, and visualizing the measurement location data of ground resistance image detection, the AI-based underground fault visualization system 100 and AI-based underground fault visualization method 200 of the present invention will provide more comprehensive and accurate visualization information of underground faults, which will be helpful for earthquake early warning, disaster prevention policy formulation, infrastructure safety assessment, and environmental disaster prevention.

[0063] Although the present invention has been disclosed above by way of embodiments, it is not intended to limit the present invention. Any person skilled in the art can make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims. [Simplified Explanation of the Diagram]

[0064] To make the above and other objects, features, advantages and embodiments of the present invention more apparent and understandable, the accompanying drawings are described as follows: Figure 1 is a block diagram of an artificial intelligence-based underground fault visualization system according to an embodiment of the present invention; Figure 2 is a flowchart of an artificial intelligence-based underground fault visualization method according to an embodiment of the present invention; and Figure 3 is a schematic diagram of a three-dimensional model according to an embodiment of the present invention.

Claims

1. An AI-based underground fault visualization system, comprising: a storage device storing at least one instruction; and a processor electrically connected to the storage device, wherein the processor is configured to access and execute the at least one instruction to: integrate geological and geophysical underground fault information using generative AI to establish a basic fault model; mark the measurement locations of resistivity image detection and combine the data of the measurement locations of the resistivity image detection with the basic fault model; invert the data of the measurement locations of the resistivity image detection to generate multiple two-dimensional cross-sectional images of the underground fault; integrate the multiple two-dimensional cross-sectional images of the underground fault into a three-dimensional model using generative AI, the three-dimensional model providing visualization of the underground fault; and analyze the changes and movement directions of the underground fault over time by combining the data of the measurement locations of the resistivity image detection at different times to predict the possible location of future earthquakes.

2. The artificial intelligence-based subsurface fault visualization system as described in claim 1, wherein the processor is configured to access and execute the at least one instruction to: identify fault structure features by comparing the data of the measurement location detected by the geoelectric resistivity image at different locations through the generative artificial intelligence, and interpret and analyze the three-dimensional fault model.

3. The AI-based subsurface fault visualization system as described in claim 1, wherein the processor is configured to access and execute the at least one instruction to: learn from hydrogeological data, historical disaster records, and public transportation data using generative AI to establish a risk prediction model, the risk prediction model being used to monitor changes in the hydrogeological data in real time to assess potential risks in advance, wherein the hydrogeological data includes soil liquefaction potential, seismic data, and data from the measurement location of the resistivity image detection.

4. The AI-based underground fault visualization system as described in claim 3, wherein the risk prediction model provides customized risk assessments based on the characteristics of different public transport routes and stations.

5. An artificial intelligence-based method for visualizing underground faults, comprising the following steps: integrating geological and geophysical underground fault information through generative artificial intelligence to establish a basic fault model; marking the measurement locations of resistivity image detection and combining the data of the measurement locations of the resistivity image detection with the basic fault model; inverting the data of the measurement locations of the resistivity image detection to generate multiple two-dimensional cross-sectional images of the underground fault; preprocessing the multiple two-dimensional cross-sectional images by at least one of the following: (a) registering the multiple two-dimensional cross-sectional images based on the measurement locations and orientations; (b) applying noise reduction; (c) applying contrast enhancement. This generative artificial intelligence integrates multiple two-dimensional cross-sectional images of the underground fault into a three-dimensional model. The integration process includes performing spatial interpolation between the multiple two-dimensional cross-sectional images. The three-dimensional model provides a visualization of the underground fault. Furthermore, by combining the data from the measurement location detected by the geoelectric resistivity image at different times, the changes and movement direction of the underground fault over time are analyzed to predict the possible location of future earthquakes.

6. The artificial intelligence-based method for visualizing underground faults as described in claim 5 further includes: using generative artificial intelligence to compare the data of the measurement location detected by the resistivity image at different locations, identifying fault structure features, and interpreting and analyzing the three-dimensional fault model.

7. The artificial intelligence-based method for visualizing subsurface faults as described in claim 5 further includes: learning from hydrogeological data, historical disaster records, and public transportation data through generative artificial intelligence to establish a risk prediction model, which is used to monitor changes in the hydrogeological data in real time to assess potential risks in advance, wherein the hydrogeological data includes soil liquefaction potential, seismic data, and data from the measurement location of the resistivity image detection.

8. The AI-based method for visualizing underground faults as described in claim 7 further includes: providing customized risk assessments based on the characteristics of different public transport routes and stations through the risk prediction model.

Images