Intelligent management method and system for mineral resources reserves based on big data

Abstract

The application discloses a mineral resource reserve intelligent management method and system based on big data, and the method comprises the following steps: obtaining a mineral resource data set, mining potential correlations between different data modes in the mineral resource data set through a self-attention mechanism, and visualizing key features through attention weights; based on the cross-modal mineral feature data, introducing a GNN graph neural network combined with InSAR technology, constructing a dynamic three-dimensional geological atlas, modeling ore bodies, faults and rock layers as graph nodes, and edges representing spatial topological relationships; generating a reserve report and a mining plan according to the ore body spatial structure atlas, and uploading the encrypted reserve report and mining plan to a blockchain network for storage by using a hash algorithm. The method can more accurately and comprehensively reflect the actual situation of mineral resources and improve resource mining efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mineral resource management, in particular to a mineral resource reserve intelligent management method and system based on big data. BACKGROUND

[0002] Traditional mineral resource data management methods mostly use manual analysis and simple statistical models, which are difficult to handle massive data, resulting in insufficient accuracy and reliability of resource reserve calculation. At the same time, geological exploration data, mining and transportation data, environmental and ecological data and other multi-source information are isolated from each other, lack effective integration and correlation analysis, and in terms of resource management, the storage and sharing of reserve reports and mining plans face data security risks, and the traceability and trustworthiness of information are low. SUMMARY

[0003] The present application is designed to solve the above problems, and provides a mineral resource reserve intelligent management method and system based on big data.

[0004] To achieve the above purpose, the technical scheme of the present application is as follows: further, in the above mineral resource reserve intelligent management method based on big data, the mineral resource reserve intelligent management method comprises the following steps:

[0005] Obtain multi-source heterogeneous data, and sequentially clean and standardize the multi-source heterogeneous data to obtain a mineral resource data set;

[0006] Mine potential correlations between different data modalities in the mineral resource data set through the self-attention mechanism of Transformer, process time series data through the spatiotemporal Transformer branch, visualize key features through attention weights, and obtain cross-modal mineral feature data;

[0007] Based on the cross-modal mineral feature data, introduce GNN graph neural network combined with InSAR technology, construct a dynamic three-dimensional geological map, model ore bodies, faults and rock layers as graph nodes, and edges represent spatial topological relationships, and output an ore body spatial structure map;

[0008] Generate a reserve report and a mining plan according to the ore body spatial structure map, encrypt the reserve report and the mining plan using a hash algorithm, and upload them to a blockchain network for storage.

[0009] Further, in the above mineral resource reserve intelligent management method based on big data, the obtaining of multi-source heterogeneous data, and the sequentially cleaning and standardizing of the multi-source heterogeneous data to obtain a mineral resource data set comprises:

[0010] Obtain geological exploration data, remote sensing data, mining and transportation data, and environmental and ecological data to obtain multi-source heterogeneous data;

[0011] According to the linear interpolation method, the missing core sample parameters in the geological exploration data in the multi-source heterogeneous data are filled according to the sample data of the adjacent depth of the same drill hole.

[0012] The missing pixel data in the remote sensing data caused by cloud cover is repaired by using a space-time interpolation algorithm combined with time series data.

[0013] The missing data in the mining and transportation data caused by equipment failure in the adjacent time period is filled by the average value, and the cleaned multi-source heterogeneous data is obtained.

[0014] Further, in the above-mentioned intelligent management method for mineral resource reserves based on big data, the multi-source heterogeneous data is obtained, and the multi-source heterogeneous data is sequentially cleaned and standardized to obtain a mineral resource data set, and the method further comprises:

[0015] The signal data in the cleaned multi-source heterogeneous data is decomposed by using a wavelet denoising algorithm, and high-frequency noise components are removed to obtain denoised multi-source heterogeneous data.

[0016] A data unit conversion dictionary is established to convert the units in different source data into a unified standard unit.

[0017] The abnormal values in the data are detected by combining the Z-score method and the IQR method, and the abnormal values are deleted to obtain the mineral resource data set.

[0018] Further, in the above-mentioned intelligent management method for mineral resource reserves based on big data, the potential correlation between different data modalities in the mineral resource data set is mined by using the self-attention mechanism of the Transformer, comprising:

[0019] A multi-modal Transformer model is constructed, and the mineral resource data set is taken as input, and the data of each modality is converted into a fixed-dimensional vector representation through an embedding layer.

[0020] The dot product operation is performed on the Query query vector of each modality, the Key key vector and the Value value vector of other modalities to obtain attention scores, and the attention weights are obtained after Softmax normalization, and the fused feature vector is obtained by weighted summation.

[0021] Further, in the above-mentioned intelligent management method for mineral resource reserves based on big data, the potential correlation between different data modalities in the mineral resource data set is mined by using the self-attention mechanism of the Transformer, comprising:

[0022] The time series data is divided by time window, and the data in each time window is input into the Transformer as a sequence, and the time self-attention mechanism is used to capture the dependency between different time points.

[0023] The key features in the time and space dimensions are displayed in the form of a heat map through attention weight visualization, and cross-modal mineral feature data is obtained.

[0024] Further, in the above-mentioned big data-based intelligent management method for mineral resource reserves, the GNN graph neural network is introduced in combination with the InSAR technology based on the cross-modal mineral feature data, including:

[0025] The ore body, fault and rock layer are modeled as graph nodes, each node containing rich attribute features, including at least the grade, reserves and geometric shape parameters of the ore body, the strike, dip angle and fault throw of the fault, and the lithology, thickness and porosity of the rock layer.

[0026] The edge represents the spatial topological relationship between nodes, including the adjacent relationship between ore bodies, the cutting relationship between ore bodies and faults, and the stratigraphic relationship between rock layers.

[0027] The surface deformation data obtained by InSAR is used as the dynamic attribute of the graph node, and the deformation information of the node is updated in real time. The InSAR data is combined with the GNN graph neural network to introduce the time dimension in the graph model, and the ore body spatial structure atlas is output.

[0028] Further, in the above-mentioned big data-based intelligent management method for mineral resource reserves, the reserves report and mining plan are generated according to the ore body spatial structure atlas, and the reserves report and mining plan are encrypted by using a hash algorithm and then uploaded to a blockchain network for storage, including:

[0029] The generated reserves report and mining plan are hashed by using a SHA-256 secure hash algorithm to generate a unique hash value, and the encrypted reserves report and mining plan are uploaded to the blockchain network in the form of a transaction, and the legality and tamper resistance of the transaction are ensured by the consensus mechanism of the blockchain.

[0030] Further, in the big data-based intelligent management system for mineral resource reserves, the mineral resource reserves intelligent management system includes the following modules:

[0031] The data acquisition and processing module is used to acquire multi-source heterogeneous data, and sequentially clean and standardize the multi-source heterogeneous data to obtain a mineral resource data set.

[0032] The cross-modal feature extraction module is configured to mine potential correlations between different data modalities in the mineral resource dataset by using a self-attention mechanism of a Transformer, process time series data by using a space-time Transformer branch, visualize key features by using attention weights, and obtain cross-modal mineral feature data.

[0033] The feature fusion processing module is configured to introduce a GNN graph neural network in combination with InSAR technology based on the cross-modal mineral feature data, construct a dynamic three-dimensional geological atlas, model ore bodies, faults and rock layers as graph nodes, and edges represent spatial topological relationships, and output an ore body spatial structure atlas.

[0034] The mineral resource management module is configured to generate a reserve report and a mining plan according to the ore body spatial structure atlas, encrypt the reserve report and the mining plan by using a hash algorithm, and upload the encrypted reserve report and the mining plan to a blockchain network for storage.

[0035] Further, in the big data-based intelligent management system for mineral resource reserves, the cross-modal feature extraction module includes the following sub-modules:

[0036] The construction sub-module is configured to construct a multi-modal Transformer model, input the mineral resource dataset, and convert data of each modality into a fixed-dimensional vector representation by using an embedding layer.

[0037] The obtaining sub-module is configured to perform dot product operations on a Query query vector of each modality, a Key key vector and a Value value vector of other modalities, obtain attention scores, obtain attention weights after Softmax normalization, and obtain a fused feature vector by using weight summation.

[0038] Further, in the big data-based intelligent management system for mineral resource reserves, the cross-modal feature extraction module includes the following sub-modules:

[0039] The input sub-module is configured to divide time series data according to time windows, input data in each time window as a sequence into a Transformer, and capture dependency relationships between different time points by using a time self-attention mechanism.

[0040] The display sub-module is configured to visualize key features in time and space dimensions in the form of a heat map by using attention weights, and obtain cross-modal mineral feature data.

[0041] The beneficial effects are that 1. it can more accurately and comprehensively reflect the actual situation of mineral resources, and provide rich and reliable data support for resource reserve evaluation and mining decision; 2. the reserve report and mining plan generated based on the ore body spatial structure map and cross-modal mineral characteristics data have higher scientificity and accuracy. The reserve calculation comprehensively considers various factors, the resource quality analysis and development and utilization condition evaluation are more comprehensive, and the mining plan can be reasonably planned according to the actual situation of the ore body and market demand, thereby improving the resource mining efficiency; 3. the data security of the reserve report and mining plan is ensured, and the data is unalterable and traceable. While protecting the commercial secrets of enterprises, the data can be trusted shared between different departments and regulatory agencies, and the transparency and collaborative efficiency of mineral resource management are improved. BRIEF DESCRIPTION OF DRAWINGS

[0042] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments with reference made to the accompanying drawings. The drawings are for purposes of illustration only and are not intended to be limiting of the present application.

[0043] Figure 1 A first embodiment schematic diagram of the mineral resource reserve intelligent management method based on big data in the embodiments of the present application;

[0044] Figure 2 A second embodiment schematic diagram of the mineral resource reserve intelligent management method based on big data in the embodiments of the present application;

[0045] Figure 3 A first embodiment schematic diagram of the mineral resource reserve intelligent management system based on big data in the embodiments of the present application. DETAILED DESCRIPTION

[0046] In order to make the objects, technical solutions and advantages of the present application clearer, the following further describes the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0047] Those skilled in the art can understand that, unless specifically stated, the singular forms "a", "an" and "the" used herein also include the plural forms. It should be further understood that the phrase "comprising" used in the specification of the present application means that the features, integers, steps, operations, elements and / or components exist, but does not exclude the existence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0048] The present application will be specifically described below in combination with the drawings, Figure 1As shown, the mineral resource reserve intelligent management method based on big data includes the following steps:

[0049] Step 101, acquiring multi-source heterogeneous data, sequentially cleaning and standardizing the multi-source heterogeneous data to obtain a mineral resource data set;

[0050] Specifically, in the present embodiment, geological exploration data, remote sensing data, mining and transportation data and environmental and ecological data are acquired to obtain multi-source heterogeneous data;

[0051] According to the linear interpolation method, the missing core sample parameters in the geological exploration data in the multi-source heterogeneous data are filled according to the sample data of the same adjacent depth of the drill hole;

[0052] The missing data of the pixel data in the remote sensing data caused by cloud cover is repaired by using the spatiotemporal interpolation algorithm combined with time series data;

[0053] The missing data of part of the period data in the mining and transportation data caused by equipment failure is filled by the average value of the adjacent time period to obtain the cleaned multi-source heterogeneous data.

[0054] Acquiring geological exploration data, remote sensing data, mining and transportation data and environmental and ecological data to obtain multi-source heterogeneous data;

[0055] According to the linear interpolation method, the missing core sample parameters in the geological exploration data in the multi-source heterogeneous data are filled according to the sample data of the same adjacent depth of the drill hole;

[0056] The missing data of the pixel data in the remote sensing data caused by cloud cover is repaired by using the spatiotemporal interpolation algorithm combined with time series data;

[0057] The missing data of part of the period data in the mining and transportation data caused by equipment failure is filled by the average value of the adjacent time period to obtain the cleaned multi-source heterogeneous data.

[0058] Specifically,

[0059] (1) Geological exploration data

[0060] Geological exploration data is obtained by various professional methods, specifically as follows:

[0061] Drilling data: drilling operations are carried out in the planned exploration area using different types of drilling machines (rotary drilling machines, percussion drilling machines), and the drilling depth is determined according to the exploration target and geological conditions.

[0062] Geophysical data

[0063] Gravity exploration: Using high-precision gravimeters to measure the intensity of the gravity field at various points on the Earth's surface, and by analyzing gravity anomalies, inferring density differences in underground geological structures and identifying potential ore bodies, faults, and other geological formations.

[0064] Magnetic exploration: Using magnetic instruments to measure the strength and distribution characteristics of the Earth's surface magnetic field, and delineating the range and geological structure of ore bodies containing magnetic minerals based on the differences in magnetic properties of different rocks and minerals.

[0065] Electrical exploration methods include resistivity methods and induced polarization methods.

[0066] Geochemical data: Soil, rock debris, and aquatic sediments are collected on the surface in grids or profiles. Analytical techniques such as atomic absorption spectrometry and inductively coupled plasma mass spectrometry are used to determine the content of chemical elements such as copper, iron, gold, and silver in the samples.

[0067] (ii) Remote sensing data

[0068] Optical remote sensing data: Using optical sensors carried by multispectral and hyperspectral satellites (Landsat, Sentinel-2, Hyperion, etc.) and UAVs, spectral data of different bands (visible light, near-infrared, short-wave infrared, etc.) of surface reflection are acquired.

[0069] Synthetic Aperture Radar (SAR) remote sensing data: Utilizing SAR satellites (Sentinel-1, Radarsat, etc.), this data acquires microwave backscattered signals from the Earth's surface. It is unaffected by weather or day / night conditions, and can penetrate clouds and some vegetation to obtain information on surface topography, geomorphology, and surface deformation. Through interferometric SAR (InSAR) technology, minute surface deformations in mining areas can be monitored, providing data support for early warning of geological disasters and monitoring of mining subsidence.

[0070] (III) Mining and Transportation Data

[0071] Mining data: Various Internet of Things (IoT) devices, such as sensors, cameras, and RFID tags, are deployed at the mining site. Pressure sensors, speed sensors, and temperature sensors are installed on mining equipment such as rock drills, crushers, and loaders.

[0072] Transportation data: GPS positioning devices and weighing sensors are installed on ore transport vehicles (trucks and trains) to obtain real-time data such as vehicle location, route, speed, and load. Weighing equipment such as weighbridges and belt scales are used at mining ports or railway freight stations to record the loading, unloading, and transportation volume of ore.

[0073] (iv) Environmental and ecological data

[0074] Atmospheric environmental data: Deploy air quality monitoring stations in the mining area and surrounding areas to monitor the concentration of pollutants such as particulate matter (PM2.5, PM10), sulfur dioxide (SO2), nitrogen oxides (NOx), carbon monoxide (CO), and ozone (O3) in the atmosphere.

[0075] Water environmental data: Install water quality monitoring sensors in rivers, lakes, and groundwater wells within the mining area to monitor water temperature, pH value, dissolved oxygen (DO), conductivity, chemical oxygen demand (COD), and ammonia nitrogen (NH3-N) in real time.

[0076] Soil environmental data: Collect soil samples in the mining area and surrounding areas according to a certain grid, and analyze indicators such as soil pH, organic matter content, heavy metal (lead, cadmium, mercury, arsenic) content, and soil fertility.

[0077] Ecological data: Obtain ecological data such as vegetation coverage, vegetation types, and biodiversity (plant species, animal population numbers) in the mining area and surrounding areas through a combination of unmanned aerial vehicle aerial photography and ground investigation.

[0078] II. Data processing

[0079] (1) Data cleaning

[0080] Data missing processing: For missing core sample parameters (porosity, permeability) in geological exploration data, use linear interpolation or cubic spline interpolation based on adjacent depth sample data in the same drill hole; for missing pixel data in remote sensing data due to cloud cover, use spatial interpolation algorithms (mean interpolation based on adjacent pixels, Kriging interpolation) combined with time series data for repair; for missing data in some time periods due to equipment failure in mining and transportation data, use the average or median of adjacent time periods for filling.

[0081] Noise data processing: For noise in geological exploration data (electromagnetic interference noise in geophysical data), use wavelet denoising algorithm for multi-scale decomposition of signals, remove high-frequency noise components, and retain effective signals; for salt and pepper noise and Gaussian noise in remote sensing image data, use spatial filtering methods such as median filtering and Gaussian filtering for smoothing; for random noise in mining equipment sensor data, use sliding window mean filtering or Kalman filtering algorithm for noise reduction.

[0082] Outlier detection and correction: Z-score method and IQR method are combined to detect outliers in various types of data. For abnormal grade values in geological exploration data, first, the data points deviating from the mean value by more than 3 times the standard deviation are preliminarily screened out by Z-score method, and then IQR method (interquartile range) is used to further confirm. For the data confirmed as outliers, according to the principles of geological statistics and adjacent sample data, the data is corrected or removed; for the abnormal load data in mining and transportation data (obviously exceeding the rated load of the vehicle), by comparing with the historical data and equipment parameters of the transport vehicle, it is judged whether it is a sensor failure or a human error. If it is a failure, the data is repaired, and if it is an error, it is removed.

[0083] (II) Data standardization

[0084] Data format unification: unified data storage format standards are established, such as for numerical data, unified floating-point or integer format is adopted, and the number of decimal places is specified; for text data, UTF-8 encoding is adopted, and the spelling and abbreviation of geological terms, mineral names, and equipment models are standardized; for time series data, ISO8601 time format ("YYYY-MM-DDHH:MM:SS") is adopted.

[0085] Data unit conversion: a data unit conversion dictionary is established to convert various units in different source data into unified standard units. For example, the pressure units (atm, bar, kPa) in geological exploration data are converted to Pa, the temperature units (℃, ℉) in remote sensing data are converted to K, and the weight units (tons, kilograms, pounds) in mining and transportation data are converted to kg.

[0086] Data normalization and standardization: different normalization methods are used for different modalities of data to make the data in the same numerical range, which is convenient for subsequent data analysis and model training. For continuous parameters (grade, porosity) in geological exploration data, the minimum-maximum normalization method is used to map the data to the [0, 1] interval; for classification data (lithology type, mineral type), One-Hot Encoding or Label Encoding method is used for conversion; for time series data, Z-score standardization method is used to make the data have zero mean and unit variance.

[0087] Step 102, the potential correlation between different data modalities in the mineral resource data set is mined through the self-attention mechanism of Transformer, the time series data is processed by the spatio-temporal Transformer branch, the key features are visualized through attention weights, and cross-modal mineral feature data is obtained;

[0088] Specifically, in this embodiment, a multi-modal Transformer model is constructed, and a mineral resource dataset is taken as input. The data of each modality is converted into a fixed-dimensional vector representation through an embedding layer.

[0089] For each modality Query query vector, other modal Key key vector and Value value vector, dot product operation is performed to obtain attention score, and after Softmax normalization, attention weight is obtained, and then weighted summation is performed through the weight to obtain a fused feature vector.

[0090] The time series data is divided according to the time window, and the data in each time window is taken as a sequence input into the Transformer. The time self-attention mechanism is used to capture the dependency between different time points.

[0091] Through attention weight visualization, the key features in the time and space dimensions are displayed in the form of a heat map to obtain cross-modal mineral feature data.

[0092] Specifically,

[0093] (I) Transformer self-attention mechanism to mine cross-modal correlation

[0094] A multi-modal Transformer model is constructed, and preprocessed geological exploration data, remote sensing data, mining and transportation data, and environmental and ecological data are taken as input. The data of each modality is first converted into a fixed-dimensional vector representation through an embedding layer. In the encoder of the Transformer, the self-attention mechanism is used to calculate the attention weight between different modal data, thereby mining the potential correlation between different data modalities. Specifically, for each modality Query query vector, other modal Key key vector and Value value vector, dot product operation is performed to obtain attention score, and after Softmax normalization, attention weight is obtained, and then weighted summation is performed through the weight to obtain a fused feature vector. Through the stacking of multiple Transformer encoders, the complex nonlinear correlation between different modal data is continuously captured.

[0095] (II) Spatiotemporal Transformer branch to process time series data

[0096] For time series data such as mining and transportation data, environmental and ecological data, a spatio-temporal Transformer branch is designed. In the time dimension, the time series data is divided into time windows, and the data in each time window is input into the Transformer as a sequence. The time self-attention mechanism is used to capture the dependency between different time points, such as the trend of mining volume over time and the seasonal fluctuation of environmental indicators. In the spatial dimension, the spatial position encoding is integrated into the Transformer model by combining the geographic coordinate information of the mining area, so that the model can consider both the time and spatial characteristics of the data. For example, for surface deformation monitoring data, the spatio-temporal Transformer can capture the deformation characteristics of different locations at different time points and their mutual influence. Through attention weight visualization technology, the key features in the time and spatial dimensions can be displayed in the form of heat maps or matrices, making it easy for analysts to intuitively identify key time points and key regions that have a significant impact on mineral resource reserve management.

[0097] Step 103, based on the cross-modal mineral feature data, introduce GNN graph neural network combined with InSAR technology to construct dynamic three-dimensional geological atlas, model ore body, fault and rock layer as graph node, edge represents spatial topological relationship, output ore body spatial structure atlas;

[0098] Specifically, in this embodiment, the ore body, fault and rock layer are modeled as graph nodes, each node contains rich attribute features, including at least the grade, reserve, geometric shape parameters of the ore body, the strike, dip angle, fault displacement of the fault, and the lithology, thickness, porosity of the rock layer;

[0099] The edge represents the spatial topological relationship between the nodes, including the adjacent relationship between the ore bodies, the cutting relationship between the ore bodies and the faults, and the stratigraphic relationship between the rock layers;

[0100] The surface deformation data obtained by InSAR is used as the dynamic attribute of the graph node, and the deformation information of the node is updated in real time. InSAR data is combined with GNN graph neural network to introduce time dimension in the graph model, and the ore body spatial structure atlas is output.

[0101] Specifically,

[0102] (I) GNN graph neural network combined with InSAR technology

[0103] Graph node definition: model the ore body, fault and rock layer as graph nodes, each node contains rich attribute features, such as the grade, reserve, geometric shape parameters (length, width, height) of the ore body, the strike, dip angle, fault displacement of the fault, and the lithology, thickness, porosity of the rock layer. These attribute features are obtained through the previous data processing and feature extraction steps.

[0104] Edge construction: Edges represent the spatial topological relationship between nodes, including the adjacency relationship between ore bodies, the cutting relationship between ore bodies and faults, the stratigraphic relationship between rock layers, etc. The connection relationship of edges is determined by calculating the spatial distance and geometric position relationship (intersection, inclusion, parallel, etc.) between nodes, and the edges are assigned corresponding weights, which can represent the inverse of spatial distance or the closeness of spatial relationship.

[0105] InSAR technology application: Use the surface deformation data obtained by InSAR as the dynamic attribute of the graph node, and update the deformation information of the node in real time. Combine InSAR data with GNN graph neural network, introduce time dimension in the graph model, and construct a dynamic graph model, which can reflect the spatial morphological changes and interaction relationship of ore bodies, faults and rock layers in the process of time passing.

[0106] (II) Construction of dynamic three-dimensional geological atlas

[0107] Based on the graph structure and node attributes constructed by GNN graph neural network, combined with the three-dimensional geological modeling technology (three-dimensional grid modeling, entity modeling) of the mining area, the spatial position and attribute characteristics of the graph node are mapped to the three-dimensional space to generate a dynamic three-dimensional geological atlas. In the three-dimensional geological atlas, ore bodies, faults and rock layers are visualized with different colors and shapes, and the connection relationship of edges clearly represents the spatial topological structure. Through real-time updating of node attribute data (change of ore body reserves in the mining process, position movement of faults caused by surface deformation), dynamic updating of three-dimensional geological atlas is realized, and the dynamic evolution process of the geological structure of the mining area is intuitively displayed. Finally, the ore body spatial structure atlas is output, which clearly presents the spatial distribution, morphological characteristics of the ore body and the relationship with the surrounding geological bodies.

[0108] Step 104, generating reserve report and mining plan according to ore body spatial structure atlas, and uploading encrypted reserve report and mining plan to blockchain network for storage.

[0109] Specifically, in this embodiment, SHA-256 secure hash algorithm is used to perform hash operation on the generated reserve report and mining plan to generate a unique hash value. The encrypted reserve report and mining plan are uploaded to the blockchain network in the form of a transaction, and the legality and tamper resistance of the transaction are ensured through the consensus mechanism of the blockchain.

[0110] Specifically,

[0111] (I) Reserve report generation

[0112] According to the ore body spatial structure atlas and the cross-modal mineral feature data obtained by mining, a detailed reserve report is generated. The content of the reserve report includes:

[0113] Mineral resource overview: The geographical location, extent, geological tectonic background, type and distribution characteristics of the mining area.

[0114] Reserve calculation: Using appropriate reserve calculation methods (geological block method, section method, three-dimensional modeling method), according to the spatial form, grade distribution and thickness of the ore body, the reserves of different types (proven reserves, controlled reserves, inferred reserves) of mineral resources are calculated, and the basis, formula and result of reserve calculation are listed.

[0115] Resource quality analysis: Evaluate the quality of mineral resources, including the grade of main useful elements, the content of impurity elements, the physical properties of ore (hardness, selectability), etc., analyze the influence of resource quality on mining and processing.

[0116] Development and utilization condition evaluation: Evaluate the mining technical conditions (ore body burial depth, inclination, hydrogeological conditions), engineering geological conditions (rock stability, fault development), environmental conditions (atmosphere, water, soil environmental quality) and economic conditions (mining cost, market demand, economic benefit prediction) of the mining area.

[0117] (2) Mining plan development

[0118] Combined with the reserve report and the actual situation of the mining area, develop a scientific and reasonable mining plan, including:

[0119] Mining sequence: Determine the mining sequence of the ore body, such as mining shallow ore bodies first and deep ore bodies later, rich ore bodies first and poor ore bodies later, to ensure the rational use of resources and mining safety.

[0120] Mining technology: Choose appropriate mining technology (open-pit mining, underground mining), according to the geological conditions of the ore body and the mining technical requirements, determine the process parameters and equipment selection of drilling, blasting, loading, transportation and other links.

[0121] Production scale: According to the reserves of mineral resources, market demand and production capacity of the enterprise, determine the reasonable annual mining capacity and production service life, ensure the feasibility and economy of the mining plan.

[0122] Safety and environmental protection measures: Develop detailed safety and environmental protection measures, including mine safety management specifications, dust and noise control schemes, wastewater and waste gas treatment measures, land reclamation and ecological restoration plans, etc., to ensure that the mining process meets the national and local safety and environmental protection standards.

[0123] (3) Hash algorithm encryption and blockchain storage

[0124] Hash algorithm encryption: SHA-256 and other secure hash algorithms are used to perform hash operations on the generated reserve reports and mining plans, generating unique hash values. The hash values are bound with the original files to ensure the integrity and tamper resistance of the files. At the same time, sensitive data (mineral resource grade, reserve specific values, mining cost, etc.) is symmetrically encrypted (AES algorithm), and the encryption key is transmitted and stored through asymmetric encryption (RSA algorithm).

[0125] Blockchain network storage: Select a suitable blockchain platform (consortium chain) for the mining industry, upload the encrypted reserve reports and mining plans to the blockchain network in the form of transactions. Each transaction contains transaction time, data digest (hash value), uploader information, etc. Through the consensus mechanism of the blockchain (PBFT, PoS), the legality and tamper resistance of the transaction are ensured. On the blockchain, once the data is stored, it can only be accessed through authorization, realizing the safe sharing and traceability of data, and providing reliable technical support for the management and supervision of mineral resources.

[0126] Its beneficial effects are: 1. It can more accurately and comprehensively reflect the actual situation of mineral resources, and provide rich and reliable data support for resource reserve evaluation and mining decision-making. 2. The reserve report and mining plan generated based on the ore body spatial structure map and cross-modal mineral feature data have higher scientificity and accuracy. The reserve calculation considers multiple factors, the resource quality analysis and development and utilization condition evaluation are more comprehensive, and the mining plan can be reasonably planned according to the actual situation of the ore body and market demand, improving the resource mining efficiency. 3. Ensure the data security, tamper resistance and traceability of the reserve report and mining plan. While protecting the commercial secrets of enterprises, it realizes the trusted sharing of data between different departments and supervision agencies, and improves the transparency and collaborative efficiency of mineral resource management.

[0127] Please refer to Figure 2 In the method for intelligent management of mineral resources reserves based on big data, multi-source heterogeneous data is obtained, and the multi-source heterogeneous data is sequentially cleaned and standardized to obtain a mineral resource data set, including the following steps:

[0128] Step 201, obtaining geological exploration data, remote sensing data, mining and transportation data, and environmental and ecological data to obtain multi-source heterogeneous data;

[0129] Step 202, according to the linear interpolation method, the missing core sample parameters in the geological exploration data in the multi-source heterogeneous data are filled according to the sample data of the same adjacent depth of the drill hole;

[0130] Step 203, using a spatiotemporal interpolation algorithm combined with time series data to repair the missing data of pixel data in the remote sensing data caused by cloud cover;

[0131] In step 204, the missing data in the mining and transportation data caused by equipment failure in the partial time period is filled by the average value of the adjacent time period, and the cleaned multi-source heterogeneous data is obtained.

[0132] The above describes the embodiments of the mineral resource reserve intelligent management method based on big data of the present application, please refer to Figure 3 In the mineral resource reserve intelligent management system based on big data, the mineral resource reserve intelligent management system comprises the following modules:

[0133] The data acquisition and processing module is used for acquiring multi-source heterogeneous data, and sequentially cleaning and standardizing the multi-source heterogeneous data to obtain a mineral resource data set.

[0134] The cross-modal feature extraction module is used for mining the potential correlation between different data modalities in the mineral resource data set through the self-attention mechanism of the Transformer, processing time series data through the space-time Transformer branch, visualizing key features through attention weights, and obtaining cross-modal mineral feature data.

[0135] The feature fusion processing module is used for introducing the GNN graph neural network combined with the InSAR technology based on the cross-modal mineral feature data, constructing a dynamic three-dimensional geological map, modeling the ore body, fault and rock layer as a graph node, and the edge representing the spatial topological relationship, and outputting the ore body spatial structure map.

[0136] The mineral resource management module is used for generating a reserve report and a mining plan according to the ore body spatial structure map, encrypting the reserve report and the mining plan by using a hash algorithm, and uploading the encrypted reserve report and the mining plan to a blockchain network for storage.

[0137] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above embodiments, and the above embodiments and descriptions in the specification are only preferred examples of the present application and are not intended to limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent management of mineral resources reserves based on big data, characterized in that, The mineral resource reserve intelligent management method comprises the following steps: Obtain multi-source heterogeneous data, and sequentially clean and standardize the multi-source heterogeneous data to obtain a mineral resource data set; Potential correlations between different data modalities in the mineral resource data set are mined through a self-attention mechanism of a Transformer, time series data are processed by a space-time Transformer branch, key features in the time and space dimensions are visualized by attention weights, and cross-modal mineral feature data are obtained; The potential correlations between different data modalities in the mineral resource data set are mined through the self-attention mechanism of the Transformer, and the method comprises the following steps: The time series data are divided into time windows, and the data in each time window is input into the Transformer as a sequence, and the time self-attention mechanism is used to capture the dependency between different time points; The key features in the time and space dimensions are displayed in the form of a heat map through attention weight visualization, and cross-modal mineral feature data are obtained; Based on the cross-modal mineral feature data, ore bodies, faults and rock layers are modeled as graph nodes, each node contains rich attribute features, at least including the grade, reserve and geometric shape parameters of the ore body, the strike, dip angle and fault throw of the fault, and the lithology, thickness and porosity of the rock layer; the edges represent the spatial topological relationship between the nodes, including the adjacent relationship between ore bodies, the cutting relationship between ore bodies and faults, and the stratigraphic relationship between rock layers; the surface deformation data obtained by InSAR are used as the dynamic attributes of the graph nodes, and the deformation information of the nodes is updated in real time; the InSAR data are combined with the GNN graph neural network, the time dimension is introduced into the graph model, and the ore body spatial structure graph is output; According to the ore body spatial structure graph, a reserve report and a mining plan are generated, the reserve report and the mining plan are encrypted by using a hash algorithm, and then uploaded to a blockchain network for storage.

2. The big data-based mineral resource reserve intelligent management method according to claim 1, characterized in that, The multi-source heterogeneous data is obtained, and the multi-source heterogeneous data is sequentially cleaned and standardized to obtain a mineral resource data set, and the method comprises the following steps: Obtain geological exploration data, remote sensing data, mining and transportation data and environmental and ecological data to obtain multi-source heterogeneous data; According to the linear interpolation method, the missing core sample parameters in the geological exploration data are filled according to the sample data of the adjacent depth of the same drill hole; The missing data of the pixel data in the remote sensing data caused by cloud cover is repaired by using a space-time interpolation algorithm combined with time series data; The missing data of some time period data in the mining and transportation data caused by equipment failure is filled by using the average value of the adjacent time period, and the cleaned multi-source heterogeneous data is obtained.

3. The big data-based mineral resource reserve intelligent management method of claim 2, wherein, The multi-source heterogeneous data is obtained, and the multi-source heterogeneous data is sequentially cleaned and standardized to obtain a mineral resource data set, and the method further comprises the following steps: The signal data in the cleaned multi-source heterogeneous data is subjected to multi-scale decomposition by using a wavelet denoising algorithm, and high-frequency noise components are removed to obtain denoised multi-source heterogeneous data; A data unit conversion dictionary is established, and the units in the data of different sources are converted into a unified standard unit; The abnormal values in the data are detected by combining the Z-score method and the IQR method, the abnormal values are deleted, and a mineral resource data set is obtained.

4. The big data-based mineral resource reserve intelligent management method of claim 1, wherein, The potential correlation between different data modalities in the mineral resource data set is mined by a self-attention mechanism of a Transformer, and the self-attention mechanism comprises the following steps: A multi-modal Transformer model is constructed, and the mineral resource data set is taken as input; data of each modality is converted into a fixed-dimensional vector representation through an embedding layer; Dot product operation is performed on the Query query vector of each modality, the Key key vector and the Value value vector of other modalities to obtain attention scores; after Softmax normalization, attention weights are obtained; and a fused feature vector is obtained by weighted summation through the weights. 5.The big data based mineral resource reserve intelligent management method according to claim 1, wherein, The reserve report and the mining plan are generated according to the ore body spatial structure graph, and the reserve report and the mining plan are encrypted by using a hash algorithm and then uploaded to a blockchain network for storage, and the method comprises the following steps: A SHA-256 secure hash algorithm is used to perform hash operation on the generated reserve report and mining plan to generate a unique hash value; the encrypted reserve report and mining plan are uploaded to the blockchain network in the form of a transaction; and the legality and tamper resistance of the transaction are ensured through the consensus mechanism of the blockchain.

6. The intelligent management system for mineral resources reserves based on big data, characterized in that, The mineral resource reserve intelligent management system comprises the following modules: A data acquisition and processing module is configured to acquire multi-source heterogeneous data, sequentially clean and standardize the multi-source heterogeneous data, and obtain a mineral resource data set; A cross-modal feature extraction module is configured to mine potential correlations between different data modalities in the mineral resource data set by using a self-attention mechanism of a Transformer, process time series data by using a space-time Transformer branch, visualize key features by using attention weights, and obtain cross-modal mineral feature data; The potential correlation between different data modalities in the mineral resource data set is mined by a self-attention mechanism of a Transformer, and the self-attention mechanism comprises the following steps: Time series data is divided into time windows, and data in each time window is taken as a sequence and input into the Transformer; and a time self-attention mechanism is used to capture the dependency relationship between different time points; Key features in the time and space dimensions are displayed in the form of a heat map by visualization through attention weights, and cross-modal mineral feature data is obtained; A feature fusion processing module is configured to model ore bodies, faults and rock layers as graph nodes based on the cross-modal mineral feature data; each node comprises rich attribute features, at least including grade, reserve and geometric morphological parameters of the ore body, strike, dip angle and fault throw of the fault, and lithology, thickness and porosity of the rock layer; edges represent spatial topological relationships between nodes, including adjacent relationships between ore bodies, cutting relationships between ore bodies and faults, and stratigraphic relationships between rock layers; surface deformation data obtained by InSAR are taken as dynamic attributes of the graph nodes, and deformation information of the nodes is updated in real time; InSAR data and a GNN graph neural network are combined, a time dimension is introduced into the graph model, and an ore body spatial structure graph is output. A mineral resource management module is configured to generate a reserve report and a mining plan according to the spatial structure map of the ore body, and to upload the encrypted reserve report and mining plan to a blockchain network for storage by using a hash algorithm.

7. The big data based mineral resource reserve intelligent management system according to claim 6, characterized in that, The cross-modal feature extraction module Comprise the following sub-modules: The construction sub-module is configured to construct a multi-modal Transformer model, take a mineral resource dataset as input, and convert data of each modality into a fixed-dimensional vector representation through an embedding layer; The obtaining sub-module is configured to perform dot product operation on a Query query vector of each modality, a Key key vector and a Value value vector of other modalities, obtain attention scores, perform Softmax normalization on the attention scores to obtain attention weights, and then obtain a fused feature vector through weighted summation of the attention weights.

8. The big data based mineral resource reserve intelligent management system according to claim 7, wherein, The cross-modal feature extraction module comprises the following sub-modules: The input sub-module is configured to divide time series data according to a time window, take data in each time window as a sequence input into a Transformer, and capture a dependency relationship between different time points by using a time self-attention mechanism; The display sub-module is configured to display key features in time and space dimensions in the form of a heat map through attention weight visualization, and obtain cross-modal mineral feature data.

Images