Geological exploration risk assessment method and device based on artificial intelligence, and medium
By using an AI-based geological exploration risk assessment method, the problems of insufficient multi-source data fusion and risk assessment feature extraction were solved, achieving higher accuracy and reliability in risk assessment, generating risk-driven exploration scheduling schemes, and promoting the intelligentization of geological exploration.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-30
- Publication Date
- 2026-04-07
AI Technical Summary
Existing geological exploration risk assessment methods have shortcomings in multi-source data fusion and risk assessment feature extraction, resulting in low accuracy and stability of assessment results.
An AI-based geological exploration risk assessment method is adopted. By collecting multi-source geological exploration data and combining it with project operation constraints, the data is uniformly organized and preprocessed to extract risk assessment features related to exploration risk decisions. These features are then fused and vectorized. Machine learning algorithms are used to generate risk score information, and stability and spatial consistency analysis are performed to ultimately generate a risk-driven exploration scheduling scheme.
It has improved the accuracy of geological exploration data processing and the reliability of assessment, enhanced the efficiency of exploration decision-making, and promoted the intelligent development of geological exploration.
Smart Images
Figure CN121810049A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of geological risk assessment technology, and in particular to a geological exploration risk assessment method, equipment and medium based on artificial intelligence. Background Technology
[0002] Geological exploration, as a crucial foundation for mineral resource development and environmental assessment, has been driven by technological innovation in recent years. With the continuous development of sensor technology, remote sensing methods, and computer technology, geological exploration is gradually moving towards digitalization and automation. For example, the combination of remote sensing data and ground sensors makes data acquisition more efficient and can encompass more diverse information. The application of big data analytics and machine learning technologies in geological data analysis has improved the ability to process large-scale geological data, helping to discover potential risks in complex geological environments and improving the accuracy and efficiency of the exploration process.
[0003] Despite advancements in data acquisition and processing, existing technologies still have limitations. Most current geological risk assessment methods rely heavily on traditional statistical analysis, which restricts the depth and breadth of data analysis. Particularly in the fusion of multi-source data and the extraction of risk assessment features, existing methods fail to fully uncover the potential correlations between data points, resulting in low accuracy and stability of the assessment results. Summary of the Invention
[0004] In view of the aforementioned existing problems, the present invention is proposed.
[0005] Therefore, this invention provides an artificial intelligence-based geological exploration risk assessment method that addresses the shortcomings in multi-source data fusion and risk assessment feature extraction.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a geological exploration risk assessment method based on artificial intelligence, which includes, Collect multi-source geological exploration data, combine it with project operation constraints, and organize it in a unified manner to output exploration assessment input data packages. Obtain standardized exploration assessment datasets through preprocessing. Risk assessment features related to exploration risk decision-making are extracted from the standardized exploration assessment dataset, and feature fusion and feature vectorization are performed on the risk assessment features to generate geological risk feature data. Machine learning risk assessment calculations are performed on geological risk characteristic data to obtain risk score information, and risk inference is performed on the risk score information to generate an initial risk distribution map; The initial risk distribution map is subjected to stability and spatial consistency analysis, and the uncertainty of the risk score is quantified and standardized to output a credible risk interval map. Risk zoning is performed on the credible risk interval map, and the target geological exploration area is divided into exploration management areas with different risk levels, generating a risk-driven exploration scheduling scheme.
[0007] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method described in this invention, the steps include: collecting multi-source geological exploration data, combining it with project operational constraints, uniformly organizing it, outputting an exploration assessment input data package, and obtaining a standardized exploration assessment dataset through preprocessing. The specific steps are as follows: Collect multi-source geological exploration data and project operation constraints, integrate them into a unified format, and output exploration assessment input data packages; The input data packets for exploration and assessment are cleaned and outliers are removed to obtain a cleaned exploration dataset. The data are then correlated to generate a correlated exploration dataset. The quality of the associated exploration dataset is verified and adjusted to obtain a standardized exploration assessment dataset.
[0008] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method of the present invention, the specific steps for extracting risk assessment features related to exploration risk decision-making from a standardized exploration assessment dataset are as follows: A risk assessment feature dataset is generated by filtering features related to exploration risk decision-making from a standardized exploration assessment dataset using a correlation analysis algorithm. The risk assessment feature dataset is expressed nonlinearly to obtain the complex relationships between features and capture the mutual influence between various risk assessment features, thereby generating an optimized feature set.
[0009] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method of the present invention, the specific steps for generating geological risk feature data by performing feature fusion and feature vectorization processing on the risk assessment features are as follows: The risk assessment feature dataset and the optimized feature set are fused to obtain multi-dimensional features, and the multi-dimensional features are reduced in dimensionality by principal component analysis to generate a low-dimensional feature dataset. Low-dimensional feature datasets are integrated and vectorized to generate geological risk feature data.
[0010] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method described in this invention, the steps of performing machine learning risk assessment calculations on geological risk characteristic data to obtain risk score information, and then performing risk inference on the risk score information to generate an initial risk distribution map are as follows: Geological risk characteristic data is input into a deep neural network and processed through multiple layers to obtain risk scoring information; Statistical inference is performed on risk scoring information to obtain risk scoring information; the risk scoring information is matched and analyzed with the geological characteristics and historical exploration data of the geological region to identify similarities and differences, and new risk scoring information is inferred based on the trends and geological characteristics of historical exploration data. The new risk score information is mapped to geographic coordinates and spatially visualized to generate an initial risk distribution map.
[0011] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method described in this invention, the steps of performing stability and spatial consistency analysis on the initial risk distribution map and performing uncertainty quantification and standardization processing on the risk score are as follows: The volatility of the initial risk distribution map is analyzed by autoregressive conditional heteroscedasticity method, and Moran's index is calculated by spatial autocorrelation analysis to assess the spatial consistency of the risk distribution and generate risk consistency data. Risk consistency data is standardized using Z-score to eliminate the influence of different data scales and is then transformed into a unified scale to generate a unified score set.
[0012] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method described in this invention, the output credible risk interval map refers to combining a unified scoring set with a standardized scoring set, assessing the uncertainty of the risk, and using spatial visualization methods to display the distribution of the risk in geographic space.
[0013] As a preferred embodiment of the artificial intelligence-based geological exploration risk assessment method described in this invention, the steps of dividing the credible risk interval map into risk zones using a risk discrimination method, and dividing the target geological exploration area into exploration management areas with different risk levels to generate a risk-driven exploration scheduling scheme are as follows: The credible risk interval map is divided into risk zones. Based on geological characteristics and risk scoring information, the geological regions are divided into different risk levels, and risk level information for each region is generated through node division rules. All areas of risk level information are evenly divided into several risk level intervals according to the frequency distribution of risk scores, thus generating the risk level range after division. Based on the risk level range after classification, geological areas with the same risk level will be merged into one exploration management area; The risk level, geological characteristics and resource requirements of each exploration management area are analyzed by a multi-objective optimization algorithm, and the priority, time arrangement and resource allocation are optimized to generate a risk-driven exploration scheduling plan.
[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the artificial intelligence-based geological exploration risk assessment method as described in the first aspect of the present invention.
[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the artificial intelligence-based geological exploration risk assessment method as described in the first aspect of the present invention.
[0016] The beneficial effects of this invention are as follows: by collecting multi-source geological exploration data and combining it with project operation constraints, the data is integrated into a unified data package, and a standardized exploration assessment dataset is generated through preprocessing using machine learning algorithms; through stability and spatial consistency analysis, a credible risk interval map is generated, and risk zoning is performed based on this map; finally, a risk-driven exploration scheduling scheme is generated through multi-objective optimization algorithms, combining geological characteristics and resource requirements; this invention effectively improves the accuracy of data processing, the reliability of assessment, and the efficiency of exploration decision-making, and promotes the intelligent development of geological exploration. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart of an artificial intelligence-based geological exploration risk assessment method.
[0019] Figure 2 This is a flowchart for feature extraction in risk assessment.
[0020] Figure 3 A flowchart for risk assessment and distribution mapping.
[0021] Figure 4 A flowchart for generating a risk-driven exploration scheduling scheme. Detailed Implementation
[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides an artificial intelligence-based geological exploration risk assessment method, including the following steps: S1. Collect multi-source geological exploration data and combine it with project operation constraints, then organize it uniformly and output exploration assessment input data packages. Obtain standardized exploration assessment datasets through preprocessing.
[0026] S1.1 Collect multi-source geological exploration data and project operation constraints, integrate them into a unified format, and output the exploration assessment input data package.
[0027] It should be noted that geological exploration data from multiple sources is collected, including geological stratigraphy, lithology, seismic wave data, and constraints related to project operation, such as time, budget, regional boundaries of geological exploration, and environmental limitations. All collected data is integrated into a unified format using specific processing methods, ensuring a consistent structure and operability. This process ensures that data from different sources can be further analyzed and processed within a unified data framework, outputting an exploration assessment input data package.
[0028] S1.2 Clean the exploration and assessment input data package and remove outliers to obtain the cleaned exploration dataset. Then, perform data association to generate the associated exploration dataset.
[0029] It should be noted that the data in the exploration assessment input data package undergoes cleaning. The data cleaning process includes identifying and removing outliers, such as measurement errors, values outside the expected range, or data deviations due to instrument malfunction. The purpose of this step is to ensure the accuracy and reliability of the exploration dataset and remove data noise that may affect subsequent analysis. After cleaning, a cleaned exploration dataset is generated to ensure the data can be effectively used in subsequent steps. Data association operations are then performed on the cleaned exploration dataset, using regression analysis algorithms to match and integrate relevant information from different data sources, ensuring the correlation between data points. Through this process, an associated exploration dataset is generated.
[0030] S1.3. Perform quality verification and adjustment on the associated exploration dataset to obtain a standardized exploration evaluation dataset.
[0031] It should be noted that quality verification is performed on the correlated exploration dataset. The quality verification process includes checking the data completeness, accuracy, and consistency to ensure that all data items conform to standards and formats and that there are no conflicts between data points. This step also includes checking for potential errors in the data, such as measurement bias, duplicate data, or missing items, to ensure the reliability and usability of the dataset. The verified data is then adjusted to correct any identified issues, such as filling in missing values, correcting data formats, or standardizing units and ranges across different data sources. This process ensures the consistency and standardization of the dataset, ultimately resulting in a standardized exploration assessment dataset.
[0032] S2. Extract risk assessment features related to exploration risk decision-making from the standardized exploration assessment dataset, and perform feature fusion and feature vectorization on the risk assessment features to generate geological risk feature data.
[0033] It should be noted that existing methods for extracting risk assessment features typically rely on traditional manual selection methods, which may lack effective fusion of multi-source data. Existing methods usually process data from different sources separately, failing to fully utilize the potential correlations between manually selected risk assessment features. Therefore, the processing of manual feature fusion and vectorization is relatively simple, unable to fully tap the potential of complex data, resulting in relatively coarse risk assessment data that cannot accurately reflect the multidimensional characteristics of geological risks.
[0034] This invention utilizes advanced machine learning methods to extract multidimensional features relevant to risk decision-making from standardized exploration and assessment datasets, and effectively integrates multi-source data through feature fusion and feature vectorization. This processing approach comprehensively considers the characteristics of different data sources and generates more accurate geological risk characteristic data. This method fully explores the correlations and potential patterns between data, improving the accuracy of risk assessment and the reliability of decision-making.
[0035] S2.1. Use correlation analysis algorithms to filter features related to exploration risk decision-making from the standardized exploration assessment dataset and generate a risk assessment feature dataset.
[0036] It should be noted that all available risk assessment feature data, including geological strata, lithology, seismic wave data, and project-related constraints (such as time, budget, and regional boundaries), are extracted from the standardized exploration assessment dataset. Correlation analysis algorithms (such as Pearson correlation coefficient and mutual information) are applied to calculate the correlation between each pair of multidimensional features, identifying risk assessment features with a strong correlation to exploration risk decisions. Using this method, multidimensional features with high correlation to exploration risk decisions are selected, while those irrelevant or with low correlation are removed. After this screening, a risk assessment feature dataset containing highly correlated features is formed. Each risk assessment feature in this dataset provides reliable information for subsequent risk assessment and decision analysis, thus contributing to the generation of geological risk feature data.
[0037] It should be noted that the expression for calculating the correlation between each pair of multidimensional features is as follows: ; in: Multidimensional features and multidimensional features The Pearson correlation coefficient between them; For the first Feature values of each sample; For the first Feature values of each sample; Multidimensional features The mean; Multidimensional features The mean; The number of samples.
[0038] S2.2. The risk assessment feature dataset is expressed nonlinearly to obtain the complex relationships between features and to capture the mutual influence between various risk assessment features, thereby generating an optimized feature set.
[0039] It should be noted that each risk assessment feature in the risk assessment feature dataset is standardized to ensure they are compared on the same scale. The correlation between each pair of risk assessment features in the dataset is calculated using common methods including Pearson correlation coefficient and mutual information. The Pearson correlation coefficient measures the linear relationship between two risk assessment features; a value closer to 1 indicates a stronger correlation. Mutual information, on the other hand, captures non-linear relationships and assesses the degree of information sharing between two risk assessment features. These methods identify risk assessment features highly relevant to exploration risk decisions and eliminate those with low correlation, thus generating a risk assessment feature dataset containing key characteristics.
[0040] S2.3. The risk assessment feature dataset and the optimized feature set are fused to obtain multi-dimensional features, and the multi-dimensional features are reduced in dimensionality by principal component analysis to generate a low-dimensional feature dataset.
[0041] It should be noted that feature fusion of the risk assessment feature dataset and the optimized feature set generates a dataset containing multi-dimensional features. These multi-dimensional feature datasets originate from different analytical perspectives and can more comprehensively describe exploration risks. Principal component analysis maps high-dimensional data to a low-dimensional space through linear transformation, preserving the main information in the data while reducing redundancy and noise. This process generates a low-dimensional feature dataset.
[0042] S2.4 Perform feature integration and vectorization on the low-dimensional feature dataset to generate geological risk feature data.
[0043] It should be noted that feature integration involves weighting and merging multi-dimensional features from different sources to ensure that each multi-dimensional feature comprehensively represents the key factors of exploration risk. The integrated multi-dimensional feature data is then vectorized, converting the data into numerical vector form that can be used as input for machine learning models. This vectorization process ensures the consistency and operability of each multi-dimensional feature in data processing. After completing the multi-dimensional feature integration and vectorization, geological risk feature data is generated.
[0044] S3. Perform machine learning risk assessment calculations on geological risk characteristic data to obtain risk score information, and infer risks from the risk score information to generate an initial risk distribution map.
[0045] S3.1 Input geological risk characteristic data into a deep neural network and obtain risk scoring information through multi-layer processing.
[0046] It should be noted that the geological risk characteristic data is input into a deep neural network for processing. The deep neural network processes the input data step by step through multiple hidden layers. Each layer performs complex feature extraction and information transformation, thereby gradually abstracting high-level features from the data. In each layer, the deep neural network transforms the high-level features nonlinearly, enabling the network to capture complex nonlinear relationships. After multi-layer processing, the deep neural network generates risk scoring information based on the feature information of the geological risk characteristic data, including geological features, historical exploration data, risk assessment features, and calculated risk levels and confidence levels.
[0047] It should also be noted that the training process of a deep neural network includes two main steps: forward propagation and back propagation. During forward propagation, the input data is processed through the various layers of the network to generate a prediction result. The back propagation step adjusts the network weights based on the error between the prediction result and the actual label (usually using a loss function such as mean squared error).
[0048] S3.2 Statistical inference is performed on the risk scoring information to obtain risk scoring information; the risk scoring information is matched and analyzed with the geological characteristics and historical exploration data of the geological area to identify similarities and differences, and new risk scoring information is inferred based on the trends and geological characteristics of historical exploration data.
[0049] It should be noted that regression analysis is used to estimate the confidence level and distribution characteristics of the risk scores. The reliability of the scores is assessed by calculating the standard error and confidence interval for each risk score. The risk score information is matched with geological features of the geological area (such as rock type and groundwater level) and historical exploration data to identify similarities and differences between the risk score information and geological features. This process helps to reveal the influence of geological features on risk scores and discover potential risk patterns. New risk score information is inferred based on trends in historical exploration data and the influence of geological features.
[0050] S3.3. Match the new risk score information with geographic coordinates and perform spatial visualization to generate an initial risk distribution map.
[0051] It should be noted that the new risk score information is mapped to geographic coordinates, that is, the risk score of each geological region is matched with its corresponding geographic coordinates (such as latitude and longitude) to ensure that each risk score accurately reflects its geographical location. Spatial visualization methods are used to combine the matched risk score information with its corresponding geographic coordinates (such as latitude and longitude), accurately pairing the risk score of each geological region with its geographical location. Mapping tools are used to mark these matched data points on a map, using visual elements such as color, size, or symbols to represent the risk scores of different regions. For example, heat maps or contour maps are used to present the spatial distribution of high-risk and low-risk areas. Through this visualization method, the preliminary risk distribution map can intuitively show the spatial distribution characteristics of geological risks, facilitating further analysis and decision-making, and generating an initial risk distribution map.
[0052] It should also be noted that spatial visualization is a method of presenting data with geographic coordinates in a two-dimensional or three-dimensional map space in a graphical way. It uses visual elements such as color coding, contour lines, and heat maps to intuitively display information such as geological risks, environmental characteristics, or statistical indicators on the spatial location, so that the relationship between geographical location and numerical information can be quickly understood and analyzed.
[0053] S4. Perform stability and spatial consistency analysis on the initial risk distribution map, and perform uncertainty quantification and standardization on the risk score to output a credible risk interval map.
[0054] It should be noted that existing methods, based on the analysis of initial risk distribution maps, typically rely on traditional statistical analysis or empirical methods, which may lack in-depth analysis of risk data and careful consideration of spatial consistency. The quantification and standardization of risk scores are usually handled through simple threshold settings or experience-based standards, which may not fully consider the dynamics and uncertainties of risk data. Therefore, the reliability of the assessment results is low, making it difficult to provide more scientifically grounded support for decision-making.
[0055] This invention employs machine learning methods to analyze the stability and spatial consistency of the initial risk distribution map, further analyzing the changing trends and spatial distribution characteristics of the risk data. For risk scoring, the invention uses uncertainty quantification and standardization methods to quantify the range of variation in risk scores, ensuring high reliability and stability, thereby outputting a credible risk interval map. This processing method significantly improves the accuracy and reliability of risk assessment, providing a more precise basis for exploration decisions.
[0056] S4.1 Analyze the volatility of the initial risk distribution map using the autoregressive conditional heteroscedasticity method, and calculate the Moran index through spatial autocorrelation analysis to assess the spatial consistency of the risk distribution and generate risk consistency data.
[0057] It should be noted that the volatility of the initial risk distribution map is analyzed using the autoregressive conditional heteroscedasticity (ARH) method. This method allows for modeling of the volatility in the risk distribution map, thereby assessing the amplitude and trend of volatility in each risk region. This process helps reveal the patterns of risk score variation, providing a foundation for subsequent risk analysis. The Moran's index is calculated using spatial autocorrelation analysis to assess the spatial consistency of the risk distribution. The Moran's index, based on its correlation with neighboring regions, determines whether spatial clustering of the risk distribution exists. A high Moran's index indicates strong spatial consistency in the risk distribution, generating risk consistency data through this analysis.
[0058] It should also be noted that the autoregressive conditional heteroscedasticity (ARH) method is used to analyze the volatility changes in time series data. The ASH method assumes that the volatility of the data is not constant but changes over time, thus dynamically capturing the volatility and risk in the data.
[0059] It should be noted that the expression for calculating the Moran index through spatial autocorrelation analysis is as follows: ; in: Calculate the Moran index for spatial autocorrelation analysis; For the region Risk score; For the region Risk score; Average risk score for all regions; For the region The set of adjacent regions.
[0060] S4.2. The risk consistency data is standardized by Z-score to eliminate the influence of different data scales and is transformed into a unified scale to generate a unified score set.
[0061] It should be noted that the risk consistency data undergoes Z-score standardization. Z-score standardization calculates the difference between each data point and the mean in the risk consistency data, and divides this difference by the standard deviation of the data, thereby eliminating the influence between different data scales. The data processed in this way has the same mean and standard deviation, allowing data points to be compared on the same scale. This standardization transforms the data into a unified scale, generating a unified scoring set.
[0062] S4.3 Combine the unified scoring set with the standardized scoring set, assess the uncertainty of risk, use spatial visualization methods to display the distribution of risk in geographic space, and output a credible risk interval map.
[0063] It should be noted that combining the unified scoring set with the standardized scoring set ensures that the data from both can be merged under the same standard. This process generates a comprehensive dataset containing all relevant scoring information by aligning and merging the data from both sets. To assess the uncertainty of risk, appropriate statistical methods, such as analysis of variance or confidence interval calculation, are used to quantify the range of variation in risk scores, reflecting the uncertainty of the risk assessment results. This process helps to understand the reliability of the assessment results and provides more information on the range of risks for decision-making. Spatial visualization methods are used to display the risk data in geographic space, mapping the risk scoring information graphically to geographic areas, clearly showing the distribution of risk in different geographical locations. Through this series of processing steps, a reliable risk interval map is output.
[0064] S5. Divide the credible risk interval map into risk zones and divide the target geological exploration area into exploration management areas with different risk levels, and generate a risk-driven exploration scheduling plan.
[0065] S5.1. Divide the credible risk interval map into risk zones, divide the geological regions into different risk levels based on geological characteristics and risk scoring information, and generate risk level information for each region through node division rules.
[0066] It should be noted that risk zoning is performed on the credible risk interval map. Based on geological characteristics (such as stratigraphic type, lithology, groundwater conditions, etc.) and risk score information, the risk levels of different geological regions are identified, and these regions are divided into different risk levels. For example, if a region has loose soil lithology, a high groundwater level, and a high risk score, it can be classified as "high-risk"; if a region has hard rock lithology, stable groundwater conditions, and a low risk score, it can be classified as "low-risk". In this way, geological regions can be divided into three levels—"high-risk," "medium-risk," and "low-risk"—based on a comprehensive analysis of geological characteristics and risk scores. This process needs to consider the relationship between geological characteristics and risk scores to ensure that the risk zoning reflects the actual geological environment. Node partitioning rules are used to generate risk level information for each region based on the distribution of risk score information and the spatial relationships of geological regions. Node partitioning rules help establish clear boundaries between different risk levels, enabling accurate and reasonable risk level classification for each geological region. Through this process, risk level information for each geological region is generated.
[0067] It should also be noted that node partitioning rules are a method for identifying and dividing different regions in data, particularly suitable for geospatial analysis. In risk assessment, node partitioning rules divide the dataset into different regions or levels based on the relationships between data points (nodes), such as geographical location, risk scores, and their changing trends. These rules are typically based on specific algorithms that can dynamically adjust partition boundaries according to data density, variation, or other characteristics.
[0068] S5.2 Divide all areas of risk level information into several risk level intervals evenly according to the frequency distribution of risk scores, and generate the risk level range after division.
[0069] It should be noted that all regions of risk level information are analyzed based on the frequency distribution of risk scores. By statistically analyzing the frequency of risk score data, the distribution of risk scores across different value ranges is identified. The regions are then evenly divided into several risk level intervals according to the frequency distribution of risk scores. Based on the characteristics of the frequency distribution, the division points are reasonably set so that each interval contains approximately the same number of regions, thereby ensuring the balance of the divided risk level intervals and generating the divided risk level range.
[0070] S5.3 Based on the risk level range after division, geological areas with the same risk level shall be merged into one exploration management area.
[0071] It should be noted that, based on the divided risk level ranges, all geological areas belonging to the same risk level are identified. These geological areas are then merged into a single exploration management area based on their shared risk level. During this process, spatial connectivity and risk level continuity between geological areas are considered to ensure consistency in risk level within each exploration management area and to reasonably reflect geological characteristics and risk distribution. The resulting exploration management area is thus generated.
[0072] S5.4 Analyze the risk level, geological characteristics and resource requirements of each exploration management area through a multi-objective optimization algorithm, and optimize the priority, time arrangement and resource allocation to generate a risk-driven exploration scheduling plan.
[0073] It should be noted that for each exploration management area, risk levels, geological characteristics, and resource requirements, including manpower, equipment, and materials, are collected and quantified. A multi-objective optimization algorithm is used to comprehensively evaluate this data, optimizing the priority, scheduling, and resource allocation for each exploration management area. During the optimization process, the complexity of risk levels and geological characteristics, as well as the finiteness of resources in each area, are taken into account to ensure that resources are allocated rationally and appropriate scheduling is developed while guaranteeing risk control and adaptability to geological characteristics, thus generating a risk-driven exploration scheduling plan.
[0074] It should also be noted that multi-objective optimization algorithms are algorithms used to solve optimization problems with multiple conflicting objectives. In practical applications, multiple objectives often cannot be optimal simultaneously; therefore, multi-objective optimization algorithms aim to find a compromise solution that balances the contradictions between different objectives.
[0075] This embodiment also provides a computer device applicable to the situation of an artificial intelligence-based geological exploration risk assessment method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the artificial intelligence-based geological exploration risk assessment method proposed in the above embodiment.
[0076] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.
[0077] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the artificial intelligence-based geological exploration risk assessment method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.
[0078] In summary, this invention achieves the following: by collecting multi-source geological exploration data and integrating it with project operational constraints into a unified data package, and preprocessing it using machine learning algorithms to generate a standardized exploration assessment dataset; by generating a credible risk interval map through stability and spatial consistency analysis, and by performing risk zoning based on this map; and finally, by combining geological characteristics and resource requirements, generating a risk-driven exploration scheduling scheme through a multi-objective optimization algorithm; this invention effectively improves the accuracy of data processing, the reliability of assessment, and the efficiency of exploration decision-making, thus promoting the intelligent development of geological exploration.
[0079] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A geological exploration risk assessment method based on artificial intelligence, characterized in that: include, Collect multi-source geological exploration data, combine it with project operation constraints, and organize it in a unified manner to output exploration assessment input data packages. Obtain standardized exploration assessment datasets through preprocessing. Risk assessment features related to exploration risk decision-making are extracted from the standardized exploration assessment dataset, and feature fusion and feature vectorization are performed on the risk assessment features to generate geological risk feature data. Machine learning risk assessment calculations are performed on geological risk characteristic data to obtain risk score information, and risk inference is performed on the risk score information to generate an initial risk distribution map; The initial risk distribution map is subjected to stability and spatial consistency analysis, and the uncertainty of the risk score is quantified and standardized to output a credible risk interval map. Risk zoning is performed on the credible risk interval map, and the target geological exploration area is divided into exploration management areas with different risk levels, generating a risk-driven exploration scheduling scheme.
2. The artificial intelligence-based geological exploration risk assessment method as described in claim 1, characterized in that: The process involves collecting multi-source geological exploration data, combining it with project operational constraints, and uniformly organizing it to output an exploration assessment input data package. A standardized exploration assessment dataset is then obtained through preprocessing. The specific steps are as follows: Collect multi-source geological exploration data and project operation constraints, integrate them into a unified format, and output exploration assessment input data packages; The input data packets for exploration and assessment are cleaned and outliers are removed to obtain a cleaned exploration dataset. The data are then correlated to generate a correlated exploration dataset. The quality of the associated exploration dataset is verified and adjusted to obtain a standardized exploration assessment dataset.
3. The artificial intelligence-based geological exploration risk assessment method as described in claim 2, characterized in that: The specific steps for extracting risk assessment features related to exploration risk decision-making from the standardized exploration assessment dataset are as follows: A risk assessment feature dataset is generated by filtering features related to exploration risk decision-making from a standardized exploration assessment dataset using a correlation analysis algorithm. The risk assessment feature dataset is expressed nonlinearly to obtain the complex relationships between features and capture the mutual influence between various risk assessment features, thereby generating an optimized feature set.
4. The artificial intelligence-based geological exploration risk assessment method as described in claim 3, characterized in that: The specific steps for performing feature fusion and feature vectorization on the risk assessment features to generate geological risk feature data are as follows. The risk assessment feature dataset and the optimized feature set are fused to obtain multi-dimensional features, and the multi-dimensional features are reduced in dimensionality by principal component analysis to generate a low-dimensional feature dataset. Low-dimensional feature datasets are integrated and vectorized to generate geological risk feature data.
5. The artificial intelligence-based geological exploration risk assessment method as described in claim 4, characterized in that: The specific steps for performing machine learning risk assessment calculations on geological risk characteristic data to obtain risk score information, and then performing risk inference based on the risk score information to generate an initial risk distribution map are as follows. Geological risk characteristic data is input into a deep neural network and processed through multiple layers to obtain risk scoring information; Statistical inference is performed on risk scoring information to obtain risk scoring information; The risk scoring information is matched and analyzed with the geological characteristics and historical exploration data of the geological area to identify similarities and differences. Based on the trends and geological characteristics of the historical exploration data, new risk scoring information is inferred. The new risk score information is mapped to geographic coordinates and spatially visualized to generate an initial risk distribution map.
6. The geological exploration risk assessment method based on artificial intelligence as described in claim 5, characterized in that: The steps for performing stability and spatial consistency analysis on the initial risk distribution map and for quantifying and standardizing the uncertainty of the risk score are as follows: The volatility of the initial risk distribution map is analyzed by autoregressive conditional heteroscedasticity method, and Moran's index is calculated by spatial autocorrelation analysis to assess the spatial consistency of the risk distribution and generate risk consistency data. Risk consistency data is standardized using Z-score to eliminate the influence of different data scales and is then transformed into a unified scale to generate a unified score set.
7. The artificial intelligence-based geological exploration risk assessment method as described in claim 6, characterized in that: The aforementioned output credible risk interval map refers to combining a unified score set with a standardized score set, assessing the uncertainty of risk, and using spatial visualization methods to display the distribution of risk in geographic space.
8. The artificial intelligence-based geological exploration risk assessment method as described in claim 7, characterized in that: The steps for performing risk zoning on the credible risk interval map and dividing the target geological exploration area into exploration management areas of different risk levels to generate a risk-driven exploration scheduling scheme are as follows: The credible risk interval map is divided into risk zones. Based on geological characteristics and risk scoring information, the geological regions are divided into different risk levels, and risk level information for each region is generated through node division rules. All areas of risk level information are evenly divided into several risk level intervals according to the frequency distribution of risk scores, thus generating the risk level range after division. Based on the risk level range after classification, geological areas with the same risk level will be merged into one exploration management area; The risk level, geological characteristics and resource requirements of each exploration management area are analyzed by a multi-objective optimization algorithm, and the priority, time arrangement and resource allocation are optimized to generate a risk-driven exploration scheduling plan.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the artificial intelligence-based geological exploration risk assessment method according to any one of claims 1 to 8.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the artificial intelligence-based geological exploration risk assessment method according to any one of claims 1 to 8.