Mine prospecting prediction system based on multi-scale big data fusion

Abstract

The invention provides a prospecting prediction system based on multi-scale big data fusion, and the system comprises a geological big data management module and a big data mining analysis module which are respectively used for carrying out the management and mining analysis of geological, geophysical, geochemical, remote sensing and mineral data; the multi-element extraction module is used for carrying out geological element extraction on the mining analysis result output by the big data mining analysis module; the multi-scale grid calculation module is used for carrying out subdivision and assignment on the geological elements through a multi-scale multi-element grid technology; the intelligent prospecting prediction module is used for compiling a result map; the intelligent prospecting evaluation module is used for compiling an optimal decision diagram; the auxiliary decision-making module generates a prospecting result report; and the prospecting dispatching command module is used for synchronously updating the data in the prospecting dispatching command module. By integrating the prospecting prediction basic data set, the knowledge base, the model base and the algorithm base, flexible scheduling and convenient use of various data, knowledge, models and algorithm tools are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of mineral resource prediction, and particularly to a prospecting prediction system based on multi-scale big data fusion. Background Art

[0002] The era of big data intelligence has led the development of geoscience technologies, promoted interdisciplinary integration and innovation, and proposed four paradigms of scientific research: experimental induction, model deduction, simulation, and intensive big data discovery. Geoscience is undergoing a major transformation from relatively independent disciplinary branches to interdisciplinary integration with information science, computer science, etc. Mineral resource prediction and evaluation based on big data and artificial intelligence technologies is one of the important directions of current international applications of geoscience big data. The development of mineral resource prediction and evaluation is tending towards comprehensive mapping, modeling and integration, fusion and evaluation of multi-factor exploration variables.

[0003] The geoscience big data required for mineral resource prediction and evaluation is a collection of geological big data at multiple scales (metallogenic belts - ore concentration areas - ore fields - ore deposits - ore bodies), multiple factors (geology, geophysics, geochemistry, remote sensing, mining engineering), and multiple dimensions (one-dimensional borehole histograms, two-dimensional section maps / mid-section maps, and three-dimensional geological body models). The development of the big data era has provided new opportunities for intelligent prospecting location prediction and quantitative prediction and evaluation of mineral resources. Geological research institutions at home and abroad are accelerating the promotion of in-depth integration and application in various fields of geoscience such as basic geological research, energy mineral investigation and evaluation, and energy mineral security. Therefore, it is necessary to develop a prospecting prediction system suitable for the fusion of geological, geophysical, geochemical, and remote sensing big data at multiple scales and multiple factors, provide a business support platform and collaborative working environment for geological prospecting personnel in mineral resource prediction and evaluation, significantly improve the efficiency of geological prospecting prediction and evaluation, form a new paradigm for geological prospecting prediction and evaluation research, and help achieve new breakthroughs in the new round of prospecting breakthrough strategic actions. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a prospecting prediction system based on multi-scale big data fusion to overcome or at least partially solve the above problems existing in the prior art.

[0005] To achieve the above object of the invention, the present invention provides a prospecting prediction system based on multi-scale big data fusion, and the system includes:

[0006] A geological big data management module for managing geological, geophysical, geochemical, remote sensing, and mineral data;

[0007] A big data mining and analysis module for mining and analyzing geological, geophysical, geochemical, remote sensing, and mineral data;

[0008] The multi-factor extraction module is used to extract geological elements from the mining and analysis results output by the big data mining and analysis module by integrating the attention mechanism;

[0009] The multi-scale grid calculation module is used to divide and assign values ​​to the geological elements of the multi-element extraction module through the multi-scale multi-element grid technology;

[0010] Intelligent prospecting prediction module, which is used to locate and quantitatively predict multi-scale mineral resources through prospecting prediction models based on the geological elements that have been segmented and assigned values, and compile result maps;

[0011] Intelligent prospecting evaluation module, used to classify the prediction area and evaluate the uncertainty and risk of the result map, and compile the optimal decision map;

[0012] The decision-making support module is used to comprehensively display the optimal decision diagram through the front-end plug-in-free visualization technology, and automatically generate a prospecting results report based on the optimal decision diagram based on natural language processing technology;

[0013] The prospecting dispatching and commanding module is used to access the optimal decision diagram and synchronously update the data in the prospecting dispatching and commanding module.

[0014] Furthermore, the geological big data management module is specifically used to execute:

[0015] By optimizing the node splitting strategy and the spatial coverage calculation method, the two-dimensional spatial data in geology, geophysics, geochemistry, remote sensing and mineral resources are processed, and the distributed tile storage of geological, geophysical, geochemical, remote sensing and mineral resources data is optimized by dividing them according to the R-tree spatial index.

[0016] Furthermore, the big data mining and analysis module is specifically used to perform:

[0017] Geological data processing, mining and analysis: selecting the initial clustering center of geological data through the clustering center initialization method according to the spatial distribution characteristics of geological data;

[0018] Geophysical data processing, mining and analysis: using deep neural networks to learn the noise and signal features in geophysical data, automatically identifying and removing noise from the data, extracting geological structure features based on graph neural networks, constructing geophysical data into graph structures, and using graph convolutional networks and graph attention networks to automatically learn the topological structure and feature information in the graph. Based on the deep learning inversion method constrained by physical information, the geophysical gravity, magnetic, and electrical physical laws and prior knowledge are integrated into the deep learning model in the form of constraints.

[0019] Geochemical data processing, mining, and analysis. An outlier processing method based on wavelet transform and Transformer. Combining the multi-resolution analysis characteristics of wavelet transform, geochemical data is decomposed into different frequency sub-bands. In different frequency sub-bands, the Transformer algorithm is used to identify weak geochemical outliers;

[0020] Remote sensing data processing, mining, and analysis. Through wavelet transform, multi-scale decomposition is performed in the spatial and spectral dimensions of remote sensing data. Combining spatio-temporal filters to remove noise, introducing time series analysis methods to perform dynamic noise suppression on hyperspectral data of continuous time, using generative adversarial networks to generate mineral spectral data sample synthetic data, training the generator and discriminator to generate mineral spectral data to expand the training sample set, and using the expanded sample set to train through classification algorithms;

[0021] Mineral data processing, mining, and analysis. Used to mine and analyze the spatial correlation relationships among ore-controlling structures, geological bodies, geophysical anomalies, geochemical anomalies, and remote sensing anomalies.

[0022] Furthermore, the multi-element extraction module is specifically used to execute:

[0023] Establish a multi-element extraction architecture based on streaming data processing, receive and process in real time the mining and analysis results output by the big data mining and analysis module, and automatically extract geological elements in the mining and analysis results through feature extraction based on the attention mechanism.

[0024] Furthermore, the multi-scale grid calculation module is specifically used to execute:

[0025] Multi-scale grid dissection. Through an adaptive multi-scale grid generation algorithm, according to the complexity and spatial distribution differences of geological elements at different scales, the geological elements are grid-dissected and the density and size of the grid are automatically adjusted. Through dynamic scale conversion and fusion technology, information is transmitted between grids of different scales;

[0026] Multi-element grid assignment. For the geological elements that have completed grid dissection, through a semantic-driven element assignment and association strategy, introducing a semantic-driven method to assign values and associate the geological elements, and through a semantic model, the semantic information of different elements is used to assign values to the geological elements that have completed grid dissection.

[0027] Furthermore, the intelligent ore prospecting prediction module is specifically used to execute:

[0028] Knowledge graph construction: Based on deep learning and knowledge graph technologies, a multi-source heterogeneous data fusion algorithm is used to automatically identify the characteristics and semantic information of different types of data in the geological big data management module. Natural language processing technology is utilized to extract prospecting-related knowledge from structured and unstructured data, and it is integrated through semantic assignment and knowledge fusion technologies to construct a geological knowledge graph. Through an adaptive knowledge reasoning and dynamic update mechanism, based on the existing knowledge in the geological knowledge graph and real-time input geological data, new prospecting knowledge and laws are automatically inferred. Reinforcement learning and uncertainty reasoning technologies are adopted to dynamically adjust the reasoning strategy, establish a dynamic update mechanism for the geological knowledge graph, and track the latest research results and exploration data in the geological field in real time to automatically update the geological knowledge graph;

[0029] Prospecting prediction model construction: By integrating the geological knowledge graph and machine learning algorithms, different machine learning models are combined, and transfer learning technology, meta-learning technology, and automated machine learning algorithms are introduced. According to the characteristics of the input geological data, problem types, and target requirements, machine learning algorithms are automatically selected, and the algorithm parameters are optimized through intelligent search algorithms to construct a prospecting prediction model;

[0030] Prospecting prediction: The geological elements that have been dissected and assigned values are input into the prospecting prediction model to conduct positioning prediction and quantitative prediction of multi-scale mineral resources, and the resulting maps are compiled.

[0031] Furthermore, the intelligent prospecting evaluation module is specifically used to execute:

[0032] Determine the prediction area as the classification basis according to the quantity and category factors of potential mineral resources within the prediction area, and divide the prediction area into regions according to the decreasing order of the ore-bearing probability;

[0033] Through multivariate statistical methods, determine the weights of geology, physics, chemistry, remote sensing, and minerals, and conduct uncertainty and risk assessment through the weights;

[0034] Through a multi-layer intelligent fusion and dynamic update algorithm, automatically identify the associations and logical relationships between different layers in the geological structure area to fuse and compile the optimal decision-making map. The different layers in the geological structure area include but are not limited to the basic layer, ore-forming element layer, prediction element layer, minimum prediction area layer, and prediction area classification and grading layer.

[0035] Furthermore, the auxiliary decision-making module is specifically used to execute:

[0036] Visualization of the optimal decision-making map: Provide a comprehensive display and query function for the infinite zoom of the optimal decision-making map by using the back-end grid data non-slice technology and the front-end non-plugin visualization technology;

[0037] Suggestions on exploration work deployment are generated through comprehensive analysis using the prospecting prediction model, thus generating the text of the suggestions for the next-step exploration work deployment;

[0038] Suggestions on prospecting verification projects are generated through comprehensive analysis using the prospecting prediction model, thus generating the text of the suggestions for the next-step prospecting verification projects;

[0039] The compilation of the prospecting results report is carried out through comprehensive analysis using the prospecting prediction model in combination with natural language processing technology, thus generating the text of the prospecting results report.

[0040] Furthermore, the prospecting scheduling and command module is specifically used to execute:

[0041] Automatically detect data changes in the auxiliary decision-making module through spatial big data caching and asynchronous transmission technology.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] The present invention proposes a prospecting prediction system based on multi-scale big data fusion. By integrating multi-source data such as geology, geophysics, geochemistry, remote sensing, and minerals, as well as unstructured information such as literature and monographs, the efficient acquisition, management, and analysis of data are realized; by combining classical algorithms with artificial intelligence technology, the geological data is deeply mined and processed, key geological elements are quickly extracted, and unified assignment and integration of data are achieved through multi-scale grid computing, effectively solving the problems of multi-source data fusion and calculation efficiency; by combining the geological knowledge graph with machine learning and deep learning algorithms, the system can accurately delineate metallogenic belts, ore concentration areas, and target areas of exploration areas, as well as optimize the deep and marginal areas of mines, and provide functions for estimating mineral resources and compiling result maps; at the same time, the present invention can compile the optimal decision-making map, has the ability to evaluate the risk of prediction results, provides scientific suggestions for exploration work deployment and prospecting verification projects, and realizes the automation of report compilation. Through integrated design, the present invention realizes the flexible scheduling and convenient use of geological data, knowledge, and models, significantly improves the efficiency and accuracy of prospecting work, provides strong support for the rapid prediction and evaluation of mineral resources, and helps to achieve the goal of a new round of prospecting breakthroughs. Brief Description of the Drawings

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0045] Figure 1Schematic diagram of the overall structure of a prospecting prediction system based on multi-scale big data fusion provided by an embodiment of the present invention. Detailed implementation manners

[0046] The present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that for the sake of description, only parts related to the present invention are shown in the drawings, rather than all the structures.

[0047] Refer to Figure 1 , this embodiment provides a prospecting prediction system based on multi-scale big data fusion, and the system includes:

[0048] A geological big data management module for managing geological, geophysical, geochemical, remote sensing, and mineral data.

[0049] The geological big data management module is specifically used to execute:

[0050] Process two-dimensional spatial data in geology, geophysics, geochemistry, remote sensing, and minerals by optimizing the node splitting strategy and the spatial coverage calculation method, and perform distributed tile storage optimization on geological, geophysical, geochemical, remote sensing, and mineral data by dividing according to the R-tree spatial index.

[0051] In this embodiment, the geological big data management module is used to manage geological spatial databases, geophysical spatial databases, geochemical spatial databases, remote sensing databases, ore deposit databases, borehole databases, unstructured database management, and business process libraries.

[0052] The spatial database proposes an R-tree spatial index technology. By optimizing the node splitting strategy (including splitting based on spatial density and data correlation algorithms) and the spatial coverage calculation method (including: adaptive boundary adjustment and elliptical coverage model), the geological big data management module can more effectively reduce data overlap, improve the index construction speed and query efficiency when processing large-scale two-dimensional spatial data, and has better performance in a high-concurrency environment; at the same time, in terms of spatial data storage and management technology, the distributed tile storage is optimized by dividing according to the R-tree spatial index strategy. Through the improved tile cutting algorithm and storage strategy, the data can be more evenly distributed on each node in a distributed environment, improving the data reading and writing performance, and at the same time supporting efficient parallel processing and data recovery.

[0053] The geological spatial database is used to manage the 1:250,000 regional geological map spatial database, the 1:200,000 regional geological map spatial database, and the 1:50,000 regional geological map spatial database, specifically including element layers such as sedimentary (volcanic) rock lithostratigraphy, intrusive rock rock ages, intrusive rock pedigrees, metamorphic rock strata (rocks), dikes (surfaces), special geological bodies, informal strata, faults, tectonic deformation zones, alteration zones (surfaces), metamorphic facies zones, mineralization zones, migmatization zones, volcanic rock facies zones, ore occurrences (points), etc.

[0054] The geophysical exploration spatial database is used to manage databases such as gravity, magnetic, electrical sounding, induced polarization electrical sounding, transient electromagnetic, magnetotelluric sounding, controlled-source audio-frequency magnetotelluric sounding, ground penetrating radar, and shallow seismic; manage the regional gravity database, including information such as measuring point data, work area information, gravity base point network data, and gravity density determination; the magnetic method database, including airborne magnetic exploration information and ground magnetic survey data information, airborne magnetic base point information, airborne magnetic exploration work area information, airborne magnetic work area range information, airborne magnetic survey data, etc.; geomagnetic base point information, ground magnetic exploration work area information, magnetic survey work area range information, geomagnetic specimen determination information, ground magnetic survey data, etc.); the ground electrical method database, containing electrical method work area information, electrical exploration survey line information, direct current resistivity method measuring point information, direct current resistivity method observation information, direct current induced polarization method measuring point information, direct current induced polarization method observation data, alternating current induced polarization method measuring point information, alternating current induced polarization method observation data, artificial source frequency sounding method measuring point information, artificial source frequency sounding method observation data, transient electromagnetic sounding method measuring point information, transient electromagnetic sounding method observation data, conventional charging method measuring point information, natural electric field method measuring point information, magnetotelluric measuring point information, magnetotelluric method observation data, etc.

[0055] The geochemical spatial database is used to manage the analysis data of water system sediments, rocks, and soil measurements and related sampling information, including geochemical exploration work area information tables, geochemical exploration work scope tables, analysis element information tables, sampling location information tables, regional water system sediment data tables, rock measurement data tables, soil measurement data tables, etc., as well as spatial calculation process data and final result data such as geochemical maps, single-element anomaly maps, combined-element anomaly maps, and comprehensive anomaly maps.

[0056] The remote sensing database is used to manage multi-source remote sensing image data (visible light, multispectral, hyperspectral, etc.) and metadata information, and at the same time manage remote sensing interpretation result data, including circular structure maps, linear structure maps, rock identification maps, mineralization maps, alteration maps, etc.

[0057] The ore deposit database is used to manage and record data such as ore deposit number, original code, ore deposit name, X coordinate, Y coordinate, ore type code, ore type name, associated ore, associated minerals, number of ore deposits, ore grade, scale, metallogenic epoch, mineralization type, tectonic position where the ore deposit is located, geological structure characteristics, ore-bearing geological body or ore-bearing horizon, wall rock alteration, industrial type, etc.

[0058] The borehole database is used to manage basic borehole information, borehole stratification information, borehole measurement data, borehole detection data, borehole analysis data, etc.

[0059] The unstructured database is used to manage various geological and mineral-related data such as literature, monographs, textbooks, achievement reports, maps, etc., providing data support for the knowledge base management module.

[0060] The business process library is used to manage the whole life cycle of geophysical, geochemical, remote sensing, and ore spatial big data management in the prospecting prediction business process, realizing the management of the initial library, process library, and result library; the business process includes data integration stage, mining and analysis stage, prospecting prediction stage, resource evaluation stage, auxiliary decision-making stage, and result output stage.

[0061] The initial library is used to realize the business process management and related data management in the data integration stage after the user selects the study area.

[0062] The process library is used to realize the business process management and related data management for the user in the mining and analysis stage, prospecting prediction stage, resource evaluation stage, and auxiliary decision-making stage.

[0063] The result library is used to realize the business process management and related data management for the user in the result output stage.

[0064] The big data mining and analysis module is used to mine and analyze geological, geophysical, geochemical, remote sensing, and mineral data. In this embodiment, it is used to perform operations such as preprocessing, spatial interpolation, anomaly extraction, information extraction, and structure identification on geological, geophysical, geochemical, remote sensing, and mineral data.

[0065] The big data mining and analysis module is specifically used to execute:

[0066] Geological data processing, mining, and analysis. According to the spatial distribution characteristics of geological data, the initial clustering center of geological data is selected by the clustering center initialization method.

[0067] Geological anomalies are the key to ore prospecting prediction and evaluation. The large-scale distributed parallel spatial computing GIS tool algorithms provided by the algorithm library are used for geological anomaly mining and analysis, including geological entropy analysis, analysis of the influence range of structures (Buffer spatial analysis), optimization of the spatial clustering algorithm through distributed parallel spatial computing, and a method for initializing the clustering center is proposed. According to the spatial distribution characteristics of the data, the initial clustering center is more reasonably selected, the number of iterations is reduced, and the clustering efficiency is improved. At the same time, the partitioning and processing methods of data in the Map and Reduce stages are optimized, so that the processing of data on the distributed cluster is more balanced, the communication overhead is reduced, and rapid clustering analysis of large-scale two-dimensional spatial data is realized; and geological spatial data editing tools, which mainly involve creating, modifying, analyzing, and displaying spatial data, mainly including point, line, and surface editing, attribute data editing, coordinate transformation, etc., and at the same time provide attribute data editing tools.

[0068] Geophysical data processing, mining and analysis. The deep neural network is used to learn the noise characteristics and signal characteristics in geophysical data, automatically identify and remove the noise in the data. Feature extraction of geological structures based on graph neural networks. The geophysical data is constructed into a graph structure, and the graph convolutional network and graph attention network are used to automatically learn the topological structure and feature information in the graph. A deep learning inversion method based on physical information constraints integrates the gravity, magnetic, and electrical physical laws and prior knowledge of geophysics into the deep learning model in the form of constraint conditions.

[0069] Using geophysical professional processing algorithms and artificial intelligence algorithms such as machine learning and deep learning provided by the algorithm library, format conversion, data cleaning, noise removal, outlier removal, missing value processing, calibration processing, data standardization, gridding, derivative calculation, preprocessing, anomaly separation, filtering, derivative calculation, analytical continuation, pole transformation, inversion, and quality assessment are carried out on gravity, magnetic, electrical, and seismic geophysical data to improve the accuracy and effectiveness of the intelligent ore prospecting prediction results. In the data preprocessing stage, a deep learning-based denoising technology is proposed. The deep neural network is used to learn the noise characteristics and signal characteristics in geophysical data. Through the comparative training of a large amount of noisy and noiseless data, the model can automatically identify and remove various types of noise in the data, such as random noise and coherent noise. Compared with traditional filtering methods, it can better retain the details and weak signals of the data.

[0070] In the feature extraction and analysis stage, a method for extracting geological structure features based on graph neural networks is proposed. Geophysical data is constructed into a graph structure, where nodes represent data points or geological bodies, and edges represent the spatial relationships or physical property relationships between data points. By using graph convolutional networks (GCNs) and graph attention networks (GATs), the topological structure and feature information in the graph are automatically learned to extract complex geological structure features such as faults and folds, improving the recognition accuracy of geological structures. In the data inversion stage, a deep learning inversion method based on physical information constraints is proposed. The gravity, magnetic, and electrical physical laws of geophysics and prior knowledge are incorporated into the deep learning model in the form of constraint conditions, enabling the model to follow physical laws while learning data features, improving the physical rationality and reliability of the inversion results, and being able to more accurately invert the underground geological structure and physical parameters.

[0071] Geochemical data processing, mining, and analysis. An outlier processing method based on wavelet transform and Transformer is proposed. Combining the multi-resolution analysis characteristics of wavelet transform, geochemical data is decomposed into different frequency sub-bands, and the Transformer algorithm is used in different frequency sub-bands to identify weak geochemical outliers.

[0072] Using geochemical professional processing algorithms and artificial intelligence algorithms such as machine learning and deep learning provided by the algorithm library, through preprocessing, anomaly analysis, and fractal processing of stream sediment, soil, and rock geochemical data, the accuracy and effectiveness of geochemical data in intelligent prospecting prediction results are improved. In the anomaly analysis stage, an outlier processing method based on wavelet transform and Transformer is proposed. Combining the multi-resolution analysis characteristics of wavelet transform, geochemical data is decomposed into different frequency sub-bands, and the Transformer algorithm is used in each sub-band to accurately identify weak geochemical outliers, providing a more accurate data basis for subsequent prospecting prediction.

[0073] Remote sensing data processing, mining, and analysis. Through multi-scale decomposition of remote sensing data in the spatial and spectral dimensions by wavelet transform, noise is removed by combining spatio-temporal filters, a time series analysis method is introduced to perform dynamic noise suppression on hyperspectral data of continuous time, and a generative adversarial network is used to generate mineral spectral data samples. By training the generator and discriminator, mineral spectral data is generated to expand the training sample set, and the expanded sample set is used for training through classification algorithms.

[0074] Using remote sensing image processing algorithms and artificial intelligence algorithms, the accuracy and effectiveness of remote sensing data in intelligent prospecting prediction results are improved through preprocessing of multi-source remote sensing data, data dimensionality reduction and feature selection, clustering analysis, image classification and target detection, spectral feature analysis, geological structure information extraction, texture and shape feature analysis, association rule mining, multi-source data fusion and analysis. In the data preprocessing stage, a multi-scale spatio-temporal joint denoising algorithm for hyperspectral data is proposed. Considering the correlation of hyperspectral data in the spatial, spectral and temporal dimensions, a multi-scale spatio-temporal joint denoising method is proposed. Wavelet transform is used for multi-scale decomposition in the spatial and spectral dimensions, and spatio-temporal filters are combined to remove noise. At the same time, time series analysis methods are introduced to suppress dynamic noise in hyperspectral data of continuous time, effectively retain useful information in the data, improve data quality, and enhance the detection ability of weak mineral spectral signals.

[0075] In the image classification and target detection stage (lithology identification, alteration identification, mineral identification, structure extraction, etc.), a hyperspectral data classification and prediction technology is proposed. The small-sample mineral classification based on generative adversarial networks is used. Aiming at the problem of few samples of certain minerals in hyperspectral data, synthetic data similar to real mineral samples are generated using generative adversarial networks (GANs); by training the generator and discriminator, the generator can generate high-quality mineral spectral data to expand the training sample set; combined with traditional classification algorithms such as support vector machines or random forests, the expanded sample set is used for training to improve the classification accuracy of small-sample mineral categories.

[0076] Mineral data processing, mining and analysis are used to mine and analyze the spatial association relationships among ore-controlling structures, geological bodies, geophysical anomalies, geochemical anomalies, and remote sensing anomalies.

[0077] Through the analysis and mining of mineral data, mainly including deposit distribution density analysis, mineralization anomaly information analysis, mineralization intensity coefficient analysis, metallogenic complexity coefficient analysis, and spatial association relationships between ore (mineralization) bodies and ore-controlling structures, geological bodies, geophysical anomalies, geochemical anomalies, and remote sensing anomalies, the advantages of various data are comprehensively utilized to improve the accuracy and reliability of prospecting information.

[0078] The multi-factor extraction module is used to extract geological elements from the mining and analysis results output by the big data mining and analysis module by integrating the attention mechanism. In this embodiment, the multi-factor extraction module proposes feature extraction based on the attention mechanism, and integrates the attention mechanism into the multi-factor extraction module, so that the module can automatically focus on the most important feature information for prospecting prediction. At the same time, a real-time dynamic extraction and update technology is proposed to establish a multi-factor extraction architecture based on streaming data processing, which can receive and process real-time data from different data sources such as geology, geophysics, geochemistry, and remote sensing in real time. When new data flows in, the characteristic variables and characteristic values ​​of the geological elements in the prospecting prediction model can be updated in real time, realizing real-time dynamic extraction and monitoring of geological variables, and timely discovering potential prospecting clues and abnormal situations.

[0079] The multi-factor extraction module is specifically used to perform:

[0080] A multi-factor extraction architecture based on streaming data processing is established to receive and process the mining and analysis results output by the big data mining and analysis module in real time, and the geological elements in the mining and analysis results are automatically extracted through feature extraction based on the attention mechanism.

[0081] In this embodiment, geological, geophysical, geochemical, remote sensing, mineral and other data are extracted according to the characteristic variables and characteristic values ​​of geological elements defined in the prospecting prediction model:

[0082] Geological element extraction focuses on studying the spatial distribution patterns and relationships between geological bodies such as structures, igneous rocks, strata or ore deposits (ore points and mineralized points), and extracting geological information that may be directly or indirectly related to regional mineralization;

[0083] Extraction of geophysical elements, including extraction and editing of gravity, magnetic, electrical and other raw geophysical data, as well as extraction of interpretation results data;

[0084] Geochemical element extraction, including raw data, geochemical anomaly maps and other information extraction functions;

[0085] Remote sensing element extraction, including annular structural map, linear structural map, rock mass, mineralized zone, alteration zone and other information extraction functions;

[0086] Mineral element extraction, including extraction based on spatial relationships, extraction based on attribute field information, etc.

[0087] The multi-scale grid computing module is used to divide and assign values ​​to the geological elements of the multi-element extraction module through the multi-scale multi-element grid technology. The multi-scale grid computing module is specifically used to perform:

[0088] Multi-scale grid meshing, through an adaptive multi-scale grid generation algorithm, meshes geological elements according to the complexity of geological elements and the differences in spatial distribution at different scales, and automatically adjusts the density and size of the grids. Through dynamic scale conversion and fusion technology, information is transmitted between grids of different scales.

[0089] In this embodiment, an adaptive multi-scale grid generation algorithm is proposed. According to the complexity of geological characteristics and the differences in spatial distribution at different scales such as metallogenic belts, ore concentration areas, ore fields, ore deposits, and ore bodies, the density and size of the grids are automatically adjusted. For example, in areas where geological characteristics change drastically and ore bodies may exist, finer grids are automatically generated; while in areas with relatively simple geological characteristics, coarser grids are used, effectively reducing the amount of calculation and data storage while ensuring data accuracy. Furthermore, a dynamic scale conversion and fusion technology is designed to achieve seamless connection and information transmission between grids of different scales. When converting from the macroscopic scale of metallogenic belts to the microscopic scale of ore bodies, the information in the large-scale grids can be accurately mapped to the small-scale grids while retaining key geological, geophysical, and other element information. Through this dynamic conversion and fusion, multi-scale grid data can better support the comprehensive analysis from macroscopic to microscopic in intelligent prospecting prediction.

[0090] Multi-element grid assignment. For the geological elements that have completed grid meshing, through a semantic-driven element assignment and association strategy, a semantic-driven method is introduced to assign values and associate the geological elements. Through a semantic model, the semantic information of different elements is used to assign values to the geological elements that have completed grid meshing. Geological elements such as geological anomalies, geophysical anomalies, geochemical anomalies, remote sensing anomalies, and ore (mineralized) body elements extracted by the multi-element extraction module are meshed and assigned values according to different scales to form a multi-channel data set, providing data preparation for the intelligent prospecting prediction module. By proposing a semantic-driven element assignment and association strategy and introducing a semantic-driven method, values are assigned and associated to multiple elements; by constructing semantic models in fields such as geology and geophysics, the semantic information of different elements is integrated into the grid assignment process. For example, for ore (mineralized) body elements, according to their semantic information such as mineral composition and genesis, semantic associations are established with geological elements such as geological anomalies and geophysical anomalies, and this association relationship is reflected in the grid assignment, thereby more deeply exploring the potential connections between multiple elements and improving the accuracy and reliability of intelligent prospecting prediction.

[0091] The intelligent prospecting prediction module is used to locate and quantitatively predict multi-scale mineral resources through a prospecting prediction model for the geological elements that have been dissected and assigned values, and compile the resulting maps. In this embodiment, the construction of the prospecting prediction model in the intelligent prospecting prediction module relies on the relevant algorithms provided by the algorithm library, the relevant geological knowledge provided by the knowledge library, and the existing models provided by the model library. A geological knowledge graph is constructed by combining the relevant algorithms in the algorithm library with a series of geological knowledge in the knowledge library, and a prospecting prediction model is constructed by combining the relevant algorithms in the algorithm library, the relevant models in the model library, and the knowledge graph.

[0092] The algorithm library is used to manage key algorithms such as GIS algorithm toolboxes, mathematical geology algorithm toolboxes, geostatistics algorithm toolboxes, classical prediction algorithm toolboxes, machine learning algorithm toolboxes, and deep learning algorithm toolboxes, and provide an extensible algorithm tool interface. A machine learning algorithm integrating geological prior knowledge is proposed, which combines the prior knowledge in the geological field with the machine learning algorithm, and introduces geological concepts, laws, and experiences as the basis for constraint conditions or feature engineering. In this way, the machine learning model can better understand and utilize the professional information in geological data, improve the accuracy and interpretability of the model, and avoid results that are contrary to geological common sense.

[0093] In the algorithm library, an adaptive machine learning algorithm selection and parameter optimization is proposed. Using meta-learning technology and automated machine learning algorithms, according to the characteristics of the input geological data, problem types, and target requirements, the most suitable machine learning algorithm is automatically adapted, and the algorithm parameters are optimized through intelligent search algorithms, greatly improving the efficiency and effect of machine learning in geological data processing, and reducing the workload and subjectivity of manual parameter tuning; the application of ensemble learning and transfer learning in geological prediction is proposed. Multiple machine learning models are combined, such as constructing an ensemble model of support vector machines, decision trees, and random forests, and the prediction results of each model are integrated through strategies such as voting and averaging to improve the stability and accuracy of the prediction. At the same time, transfer learning technology is introduced to transfer the model knowledge trained in other similar geological regions or related fields to the new geological data scenario, accelerating the model training speed and improving the adaptability of the model to new data.

[0094] Aiming at the problems of limited sample size and data imbalance commonly existing in geological data, a deep learning model based on data augmentation is proposed. By performing data augmentation operations such as rotation, scaling, translation, and adding noise on the prediction element data in the prospecting prediction model, the scale of the dataset is expanded and the diversity of the data is increased. At the same time, in the model structure design, techniques such as spatial attention mechanisms and multi-element (geological, physical, chemical, remote sensing, mineral, etc. features) geological attention mechanisms are adopted, enabling the model to pay more attention to the key spatial features and geological features in the data, and improving the learning ability and generalization ability of the model for geological data.

[0095] By integrating the feature representations of two networks, the joint modeling and analysis of multi-modal geological data are realized. A multi-modal deep learning algorithm fusion method is proposed to fuse different types of deep learning algorithms to construct a multi-modal deep learning model. For example, a convolutional neural network (CNN) is used to extract the spatial features of geological image data, and a recurrent neural network (RNN) or a long short-term memory network (LSTM) is used to process geological process simulation data to more comprehensively extract information from geological data and improve the understanding and prediction ability of geological phenomena.

[0096] To explain the basis and process of prediction and help geologists understand the decision-making mechanism of the model, the application of interpretable deep learning algorithms in the geological field is proposed, such as interpretable methods based on rule extraction, feature importance analysis, etc., which are applied to the analysis and prediction of geological data. This enables the model to not only give accurate prediction results but also enhance the trust in the deep learning model, promoting the wide application of deep learning in the geological field.

[0097] The knowledge base is used to manage the prospecting knowledge graph and the large model system, realizing functions such as prospecting knowledge Q&A, summarization, and report writing. Facing the widely sourced and complex-structured data in the geological field, a knowledge extraction and integration algorithm for multi-source heterogeneous data fusion and a multi-source heterogeneous data fusion algorithm based on deep learning and knowledge graph technology are proposed. These algorithms can automatically identify the characteristics and semantic information of different types of data (such as geological reports, exploration data, remote sensing images, etc.), extract prospecting-related knowledge from structured and unstructured data using natural language processing technology, and integrate this knowledge into a unified knowledge graph through semantic alignment and knowledge fusion technology, effectively solving the problems of information inconsistency and incompleteness existing in traditional methods when dealing with multi-source data.

[0098] Using an adaptive knowledge reasoning and dynamic update mechanism, an adaptive knowledge reasoning engine is developed. It can automatically infer new prospecting knowledge and rules based on the existing knowledge in the knowledge graph and the real-time input geological data. By adopting reinforcement learning and uncertainty reasoning techniques, it can dynamically adjust the reasoning strategy to improve the accuracy and reliability of reasoning. At the same time, a dynamic update mechanism for the knowledge graph is established to track the latest research results and exploration data in the geological field in real time and automatically update the knowledge graph to ensure the timeliness and accuracy of knowledge.

[0099] Using large prediction model technology, on the basis of the basic large model, overlaying the prospecting geological and mineral datasets, after pre-training, fine-tuning, and alignment, a large prospecting knowledge model is formed to achieve functions such as prospecting knowledge Q&A, reasoning on metallogenic geological background research, reasoning on metallogenic regularity research, reasoning on ore deposit model research, reasoning on prospecting model research, establishment of prospecting prediction models, summary of prospecting prediction results, compilation of suggestions for the next step of geological work, and compilation of prospecting prediction result reports;

[0100] In the pre-training stage, a pre-training technology for large prospecting knowledge models with multi-modal data fusion is proposed. The multi-modal data fusion technology is innovatively introduced, and various types of data such as spatial database data, text reports, and pictures in the fields of geology, geophysics, geochemistry, remote sensing, and minerals are input into the large model for joint training. By designing a special multi-modal fusion layer and attention mechanism, the model can fully learn the associations and complementary information between different modal data, improving the model's understanding and expression ability of prospecting knowledge. Compared with traditional single-modal pre-training methods, this technology enables the model to better process complex geological data and enhances the model's generalization ability and accuracy.

[0101] Aiming at the problems of relatively small data volume and high annotation cost in the prospecting field, a model fine-tuning method based on transfer learning and domain adaptation and a model fine-tuning method based on transfer learning and domain adaptation are proposed. First, the basic large model is pre-trained with large-scale general domain data to learn general language and knowledge representations; then, fine-tuning is carried out on the prospecting geological and mineral datasets. By introducing domain adaptation technology, the model can quickly adapt to the characteristics and requirements of the prospecting field, reducing the dependence on a large amount of annotated data; a knowledge graph and a logical reasoning module are introduced. By associating and reasoning the prediction results of deep learning with the prior knowledge in the knowledge graph, interpretable prospecting conclusions and bases are generated; at the same time, a strategy for dynamically adjusting fine-tuning parameters is designed to automatically optimize the fine-tuning process according to the data distribution and model performance, improving the fine-tuning effect of the model.

[0102] The model library is used to manage expert prospecting prediction models and AI prospecting prediction models: The expert prospecting prediction model management module is a prospecting prediction model summarized by geological experts based on the research and understanding of metallogenic regularity; The AI prospecting prediction model management module is a prospecting prediction model summarized by the large model system after pre-training based on a large amount of geological data (knowledge-driven). The intelligent prospecting prediction module summarizes by extracting and fusing multi-element spatial data of typical minerals (data-driven) combined with the regional metallogenic geological background and regional metallogenic regularity to generate a prospecting prediction model, which is the result of the integration of knowledge-driven and data-driven.

[0103] First, the classification and hierarchical management technology is proposed. According to the source and nature of the prospecting prediction model, the model library is divided into the expert prospecting prediction model management area and the AI prospecting prediction model management area. The models in different areas are hierarchically managed according to factors such as their complexity and scope of application. For the expert prospecting prediction models, they are further subdivided into levels according to different metallogenic types (such as magmatic type, sedimentary type, etc.); for the AI prospecting prediction models, they are hierarchically divided according to the types of geological data used in pre-training (such as geophysical data, geochemical data, etc.). In terms of storage structure, a tree directory structure is adopted for organization, which is convenient for quickly locating and managing different types of models.

[0104] Secondly, the knowledge graph association technology is proposed. For the expert prospecting prediction models, various types of knowledge in the study of metallogenic laws (such as geological structures, rock characteristics, mineralization signs, etc.) are associated with the models using the knowledge graph. For the AI prospecting prediction models, the knowledge nodes in the massive geological data involved in model pre-training are connected to the models through the knowledge graph. At the same time, combined with the knowledge related to the regional metallogenic geological background and regional metallogenic laws in the massive database, a more comprehensive knowledge graph association is supplemented and constructed. When querying or calling a model, through the relationship network of the knowledge graph, the knowledge and information related to the model can be quickly obtained to assist in the selection and application of the model.

[0105] Finally, the dynamic update and adaptive adjustment technology is proposed. With the acquisition of new geological data, the in-depth study of metallogenic laws, and the optimization of AI algorithms, various prospecting prediction models need to be continuously updated. A dynamic update mechanism is established, and when new data or knowledge appears, the model update process is automatically triggered. For the AI prospecting prediction models, online learning technology is used to enable them to adjust model parameters in real time according to new data; for the expert prospecting prediction models, a visual editing interface is provided to facilitate geological experts to modify and improve the models according to the latest research results. At the same time, the model library can adaptively adjust the application strategies and parameter settings of the models according to information such as changes in geological conditions in different regions and feedback from prospecting practices.

[0106] The intelligent prospecting prediction module is specifically used to execute:

[0107] Knowledge graph construction: Based on deep learning and knowledge graph technologies, a multi-source heterogeneous data fusion algorithm is used to automatically identify the characteristics and semantic information of different types of data in the geological big data management module. Natural language processing technology is utilized to extract prospecting-related knowledge from structured and unstructured data, and it is integrated through semantic assignment and knowledge fusion technologies to construct a geological knowledge graph. Through an adaptive knowledge reasoning and dynamic update mechanism, based on the existing knowledge in the geological knowledge graph and real-time input geological data, new prospecting knowledge and laws are automatically inferred. Reinforcement learning and uncertainty reasoning technologies are adopted to dynamically adjust the reasoning strategy, establish a dynamic update mechanism for the geological knowledge graph, and track the latest research results and exploration data in the geological field in real time to automatically update the geological knowledge graph.

[0108] Prospecting prediction model construction: By integrating the geological knowledge graph and machine learning algorithms, different machine learning models are combined, and transfer learning technology, meta-learning technology, and automated machine learning algorithms are introduced. According to the characteristics, problem types, and target requirements of the input geological data, machine learning algorithms are automatically selected, and the algorithm parameters are optimized through intelligent search algorithms to construct a prospecting prediction model.

[0109] Prospecting prediction: The geological elements that have been dissected and assigned values are input into the prospecting prediction model to conduct positioning prediction and quantitative prediction of multi-scale mineral resources, and the resulting maps are compiled.

[0110] In this embodiment, the intelligent prospecting prediction module uses classical prediction, machine learning, and deep learning algorithms in the algorithm library, as well as multi-element geo-physical, chemical, and remote sensing big data generated by the model library and multi-scale grid computing mode for large-scale data fusion and calculation, realizing functions such as delineating prospective areas in metallogenic belts, delineating target areas in ore concentration areas, delineating target areas in exploration areas, and optimizing deep and marginal blocks of mines. At the same time, it provides the function of estimating the resource volume of mineral resources, and then compiles the resulting maps, including prediction area maps, prediction target area maps, resource volume maps, etc.

[0111] Classical prediction algorithm module: The weight-of-evidence method, information content method, feature analysis method, and fuzzy weight-of-evidence method are adopted.

[0112] Weight-of-evidence algorithm: Conduct prospecting prediction according to the business process of the weight-of-evidence method for mineral prediction. First, use the prediction elements defined by the expert prospecting prediction model or AI prospecting prediction model in the model library module to construct and preprocess the evidence layer (determine the prediction elements in the prospecting prediction model and perform binary processing of the data), then calculate the weight of evidence (count the number of units of each variable and the number of ore-bearing units, calculate the positive and negative weights, and calculate the contrast coefficient), conduct conditional independence tests again, then calculate the posterior probability and delineate the prospective area, and finally conduct model testing and evaluation.

[0113] Information quantity method algorithm: Conduct prospecting prediction according to the business process of the information quantity method. First, use the prediction elements defined by the expert prospecting prediction model or the AI prospecting prediction model in the model library module to determine and assign prospecting signs (prospecting sign identification, assignment calculation). Secondly, conduct comprehensive information quantity calculation. Then, conduct prospecting prediction to determine the favorable ore-forming areas. Finally, conduct model verification and evaluation.

[0114] Feature analysis algorithm: Conduct prospecting prediction according to the business process of the feature analysis method. First, use the prediction elements defined by the expert prospecting prediction model or the AI prospecting prediction model in the model library module to conduct feature analysis and variable extraction. Secondly, conduct preprocessing of feature variables using data standardization. Then, establish a prediction model, conduct model training and model evaluation. Then, delineate the prospective area and conduct result verification and optimization.

[0115] Estimation of mineral resource reserves: According to the delineated prospective area, use the volume method to estimate the reserves based on the experience of prospecting experts and the data parameters of information technology means.

[0116] In this embodiment, an integrated learning framework for multi-algorithm fusion is constructed, which organically combines classical prediction algorithms (such as the weight of evidence method, information quantity method, feature analysis method, fuzzy weight of evidence method) with machine learning and deep learning algorithms. In this framework, different algorithms can play their respective advantages, and the results are fused through voting mechanisms, weighted averages, etc. For example, classical prediction algorithms are based on geostatistical principles and can provide relatively stable basic prediction results; while machine learning algorithms (such as random forest, support vector machine) have advantages in dealing with non-linear relationships and complex data patterns; deep learning algorithms (such as convolutional neural network, recurrent neural network) are good at mining deep features in data.

[0117] Through the integrated learning framework, the prediction results of these algorithms are fused, which can effectively improve the accuracy and reliability of prospecting prediction. In the link of estimating the mineral resource reserves, an uncertainty quantification method is introduced. The traditional volume method for estimating reserves often ignores the uncertainty of geological data and the error factors in the estimation process. The new method conducts uncertainty analysis on various parameters (such as the boundary, thickness, grade, etc. of the ore body) in the estimation process through technologies such as Monte Carlo simulation and Bayesian inference. For example, use Monte Carlo simulation to randomly generate parameter values multiple times, calculate the reserves under different parameter combinations, so as to obtain the probability distribution range of the reserves, providing more comprehensive and accurate information for the evaluation of mineral resources, enabling decision-makers to fully understand the uncertainty risks of reserve estimation.

[0118] Intelligent prospecting evaluation module, used to divide the prediction area level of the achievement drawings and conduct uncertainty and risk evaluation, and compile the optimal decision-making drawings.

[0119] The intelligent prospecting evaluation module is specifically used to execute:

[0120] Determine the prediction area as the classification basis according to the quantity and category factors of potential mineral resources within the prediction area, and divide the prediction area into regions according to the ore-bearing probability from high to low;

[0121] Through multivariate statistical methods, determine the weights of geology, physics, chemistry, remote sensing, and minerals, and conduct uncertainty and risk evaluations through the weights;

[0122] Through the multi-layer intelligent fusion and dynamic update algorithm, automatically identify the associations and logical relationships between different layers in the geological structure area, so as to fuse and compile the optimal decision-making map. The different layers in the geological structure area include but are not limited to the basic layer, ore-forming element layer, prediction element layer, minimum prediction area layer, and prediction area classification and grading layer.

[0123] In this embodiment, the prediction area is determined as the classification basis according to factors such as the quantity and category of potential mineral resources within the prediction area, and the prediction area is optimized and classified. The prediction area is divided into three categories, A, B, and C, according to the ore-bearing probability from high to low. Then, uncertainty and risk evaluations are carried out, comprehensively considering various prediction factors of geology, geophysics, geochemistry, and remote sensing, quantifying these uncertainties and risks, providing a scientific basis for mineral development decisions, and ensuring the rationality of investment and the sustainability of the project. For the optimization and classification of mineral prediction areas, a multi-dimensional dynamic prediction area classification and optimization algorithm is proposed. Traditional prediction area classification and optimization mainly rely on single or a few factors, while the present invention innovatively integrates multi-source data of geology, geophysics, geochemistry, and remote sensing to construct a comprehensive and detailed comprehensive evaluation index system. For example, not only the quantity and category of potential mineral resources in the prediction area are considered, but also the geophysical anomaly characteristics (such as the intensity and shape of gravity and magnetic anomalies), the combination and distribution law of geochemical elements, the ore-controlling effect of geological structures, and the geological characteristics reflected in remote sensing images are deeply analyzed.

[0124] Through multivariate statistical methods such as principal component analysis and analytic hierarchy process, determine the weights of each index to achieve a comprehensive and objective evaluation of the prospecting potential of the prediction area. A dynamic adaptive prediction area classification and optimization model is developed, which can automatically adjust the classification and optimization results according to newly acquired real-time data (such as newly added exploration data, research results, etc.). The online learning algorithm in machine learning is used to enable the prospecting prediction model to continuously learn the characteristics and laws in the new data and update the evaluation of the prediction area. For example, when a special geological structure or abnormal geochemical element enrichment phenomenon is newly discovered in a certain prediction area, the prospecting prediction model can quickly take it into account, re-evaluate the prospecting potential of the prediction area, adjust its classification and optimization level, and ensure that the prospecting evaluation is always based on the latest and most accurate information.

[0125] Compiling the optimal decision-making map: The mineral prediction result map of the prediction area is generated through the thematic base map of geological structures in the prediction work area - metallogenic element map - prediction element map - mineral prediction map. The layers include the basic geological structure layer, metallogenic element layer, prediction element layer, minimum prediction area layer, and prediction area classification and grading layer. When compiling the optimal decision-making map, a multi-layer intelligent fusion and dynamic update algorithm is proposed, which can automatically identify the associations and logical relationships between different layers such as the basic geological structure layer, metallogenic element layer, prediction element layer, minimum prediction area layer, and prediction area classification and grading layer, and achieve efficient and accurate fusion. For example, through semantic matching and spatial position association, the geological structure information and metallogenic element information are organically combined, enabling the decision-making map to more comprehensively reflect the relationship between geological conditions and prospecting potential. At the same time, the algorithm supports real-time dynamic updates. When new geological data or research results are available, it can quickly update the corresponding layers and automatically re-fuse to generate the latest optimal decision-making map.

[0126] The auxiliary decision-making module is used to comprehensively display the optimal decision-making map through front-end plugin-free visualization technology, and automatically generate a prospecting result report based on natural language processing technology according to the optimal decision-making map. Specifically, the auxiliary decision-making module is used to execute:

[0127] Visualization of the optimal decision-making map, using back-end grid data non-slice technology and front-end plugin-free visualization technology to provide the optimal decision-making map with infinite zoom for comprehensive display and query functions;

[0128] Suggestions for exploration work deployment, through comprehensive analysis of the prospecting prediction model, to generate text for suggestions on the next-step exploration work deployment;

[0129] Suggestions for prospecting verification projects, through comprehensive analysis of the prospecting prediction model, to generate text for suggestions on the next-step prospecting verification projects;

[0130] Compiling the prospecting result report, through comprehensive analysis of the prospecting prediction model combined with natural language processing technology, to generate text for the prospecting result report.

[0131] In this embodiment, based on the natural language processing ability of the prospecting prediction model, intelligent text generation of the prospecting result report is realized. The prospecting prediction model can automatically organize language to generate a report text with clear structure and complete content according to the results of big data analysis and the conclusions of spatial comprehensive analysis and evaluation. During the text generation process, the prospecting prediction model follows the norms and report templates of the geological and mining industry to accurately express information such as the purpose, method, results, and conclusions of prospecting work. For example, when compiling the prospecting prediction result report, the prospecting prediction model can elaborate on the prospecting potential, existing risks, and suggestions for the next-step work in detail based on the analysis results of the predicted target areas, providing a comprehensive and accurate reference basis for decision-makers.

[0132] Adopt the technology of no slicing for backend grid data and the technology of no plugin visualization for the front end to provide the comprehensive display and query function of infinite zooming for prospecting areas (target areas), and improve the loading speed and effect of visualization. Adopt the large model technology, and through pre-training the large model of the vertical industry, and at the same time combine the results generated by the spatial comprehensive analysis and evaluation in the intelligent prospecting prediction business process within this system, to generate the text for the next exploration work deployment, the text for the suggestion of prospecting verification projects, and the text for the prospecting result report.

[0133] The prospecting scheduling and command module is used to access the optimal decision-making diagram and synchronously update the data in the prospecting scheduling and command module. The prospecting scheduling and command module is specifically used to execute:

[0134] Automatically detect the data changes in the auxiliary decision-making module through the spatial big data caching and asynchronous transmission technology.

[0135] In this embodiment, in order to ensure that the prospecting scheduling and command module can obtain the latest prospecting data, a data real-time update and synchronization mechanism is developed. By establishing a real-time data connection with the auxiliary decision-making module, the real-time collection and transmission of the optimal decision-making diagram data are realized. When the data in the auxiliary decision-making module changes, the system can automatically detect and update in time to ensure the consistency and timeliness of the data. At the same time, adopt the spatial big data caching and asynchronous transmission technology to improve the efficiency and stability of data update, and avoid affecting the normal operation of the system due to data transmission delay or interruption.

[0136] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A prospecting prediction system based on multi-scale big data fusion, characterized in that: The system comprises: Geological big data management module, used to manage geological, geophysical, geochemical, remote sensing and mineral data; Big data mining and analysis module, used for mining and analysis of geological, geophysical, geochemical, remote sensing and mineral data; The multi-factor extraction module is used to extract geological elements from the mining and analysis results output by the big data mining and analysis module by integrating the attention mechanism; The multi-scale grid calculation module is used to divide and assign values ​​to the geological elements of the multi-element extraction module through the multi-scale multi-element grid technology; Intelligent prospecting prediction module, which is used to locate and quantitatively predict multi-scale mineral resources through prospecting prediction models based on the geological elements that have been segmented and assigned values, and compile result maps; Intelligent prospecting evaluation module, used to classify the prediction area and evaluate the uncertainty and risk of the result map, and compile the optimal decision map; The decision-making support module is used to comprehensively display the optimal decision diagram through the front-end plug-in-free visualization technology, and automatically generate a prospecting results report based on the optimal decision diagram based on natural language processing technology; The prospecting dispatching and commanding module is used to access the optimal decision diagram and synchronously update the data in the prospecting dispatching and commanding module.

2. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The geological big data management module is specifically used to execute: By optimizing the node splitting strategy and the spatial coverage calculation method, the two-dimensional spatial data in geology, geophysics, geochemistry, remote sensing and mineral resources are processed, and the distributed tile storage of geological, geophysical, geochemical, remote sensing and mineral resources data is optimized by dividing them according to the R-tree spatial index.

3. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The big data mining and analysis module is specifically used to perform: Geological data processing, mining and analysis: selecting the initial clustering center of geological data through the clustering center initialization method according to the spatial distribution characteristics of geological data; Geophysical data processing, mining and analysis: using deep neural networks to learn the noise and signal features in geophysical data, automatically identifying and removing noise from the data, extracting geological structure features based on graph neural networks, constructing geophysical data into graph structures, and using graph convolutional networks and graph attention networks to automatically learn the topological structure and feature information in the graph. Based on the deep learning inversion method constrained by physical information, the geophysical gravity, magnetic, and electrical physical laws and prior knowledge are integrated into the deep learning model in the form of constraints. Geochemical data processing, mining and analysis, based on the outlier processing method of wavelet transform and Transformer, combined with the multi-resolution analysis characteristics of wavelet transform, the geochemical data is decomposed into different frequency sub-bands, and the Transformer algorithm is used in different frequency sub-bands to identify geochemical weak outliers; Remote sensing data processing, mining and analysis: multi-scale decomposition of remote sensing data in spatial and spectral dimensions is performed through wavelet transform, and noise is removed by combining spatiotemporal filters. Time series analysis methods are introduced to dynamically suppress noise in continuous-time hyperspectral data. Generate synthetic data of mineral spectral data samples using generative adversarial networks. By training generators and discriminators, mineral spectral data is generated to expand the training sample set, and the expanded sample set is used for training through classification algorithms. Mineral data processing, mining and analysis are used to mine and analyze the spatial correlation between ore-controlling structures, geological bodies, geophysical anomalies, geochemical anomalies and remote sensing anomalies.

4. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The multi-factor extraction module is specifically used to perform: A multi-factor extraction architecture based on streaming data processing is established to receive and process the mining and analysis results output by the big data mining and analysis module in real time, and the geological elements in the mining and analysis results are automatically extracted through feature extraction based on the attention mechanism.

5. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The multi-scale grid computing module is specifically used to perform: Multi-scale grid generation: Adaptive multi-scale grid generation algorithm is used to grid geological elements and automatically adjust the density and size of grids according to the complexity and spatial distribution of geological elements at different scales. Dynamic scale conversion and fusion technology is used to transfer information between grids of different scales. Multi-element grid assignment, the geological elements that have completed grid division are assigned values ​​and associated with each other through semantic-driven element assignment and association strategies, and the semantic-driven method is introduced to assign values ​​and associate geological elements. Through the semantic model, the semantic information of different elements is assigned to the geological elements that have completed grid division.

6. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The intelligent prospecting prediction module is specifically used to perform: Knowledge graph construction, based on the multi-source heterogeneous data fusion algorithm of deep learning and knowledge graph technology, automatically identifies the characteristics and semantic information of different types of data in the geological big data management module, uses natural language processing technology to extract prospecting-related knowledge from structured and unstructured data, and integrates them through semantic assignment and knowledge fusion technology to build a geological knowledge graph. Through adaptive knowledge reasoning and dynamic update mechanism, according to the existing knowledge in the geological knowledge graph and the real-time input geological data, new prospecting knowledge and laws are automatically inferred. Reinforcement learning and uncertainty reasoning technology are used to dynamically adjust the reasoning strategy, establish a dynamic update mechanism for the geological knowledge graph, track the latest research results and exploration data in the geological field in real time, and automatically update the geological knowledge graph; The construction of the prospecting prediction model combines different machine learning models by integrating geological knowledge graphs and machine learning algorithms, introduces transfer learning technology, meta-learning technology and automatic machine learning algorithms, automatically selects machine learning algorithms according to the characteristics of input geological data, problem types and target requirements, and optimizes algorithm parameters through intelligent search algorithms to build a prospecting prediction model; Prospecting prediction: input the geological elements that have been segmented and assigned values ​​into the prospecting prediction model, conduct positioning prediction and quantitative prediction of multi-scale mineral resources, and compile the results map.

7. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The intelligent prospecting evaluation module is specifically used to perform: Determine the prediction area based on the quantity and category of potential mineral resources within the prediction area as the classification basis, and divide the prediction area into regions from high to low probability of mineralization; Through multivariate statistical methods, the weights of geology, physics, chemistry, remote sensing and minerals are determined, and the uncertainty and risk are evaluated through the weights; Through multi-layer intelligent fusion and dynamic update algorithm, the association and logical relationship between different layers in the geological structure area are automatically identified to integrate and compile the optimal decision map. The different layers in the geological structure area include but are not limited to basic layers, mineralization element layers, prediction element layers, minimum prediction area layers and prediction area classification and grading layers.

8. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The auxiliary decision module is specifically used to execute: The optimal decision diagram visualization uses the back-end grid data non-slicing technology and the front-end plug-in-free visualization technology to provide the optimal decision diagram with infinite zooming for comprehensive display and query functions; Proposal for deployment of exploration work: through comprehensive analysis of prospecting prediction models, a text of proposals for deployment of the next exploration work is generated; Prospecting verification project proposal, through comprehensive analysis of prospecting prediction model, to generate the text of the next prospecting verification project proposal; The prospecting results report is prepared by conducting a comprehensive analysis through the prospecting prediction model combined with natural language processing technology to generate the text of the prospecting results report.

9. The mineral prospecting prediction system based on multi-scale big data fusion according to claim 1 is characterized in that: The ore prospecting dispatching and commanding module is specifically used to execute: Data changes in the auxiliary decision-making module are automatically detected through spatial big data caching and asynchronous transmission technology.

Images