Intelligent mining quantity accurate calculation system based on three-dimensional geologic model

Abstract

The invention relates to the technical field of accurate calculation of mining quantity, in particular to an intelligent accurate calculation system of mining quantity based on a three-dimensional geologic model, which comprises a multi-source heterogeneous data acquisition module used for acquiring and preprocessing various data such as geologic structure and mineral distribution; the dynamic three-dimensional geologic modeling module constructs a digital geologic model; the precise reserve calculation module calculates exploitable reserves; the intelligent mining planning module is used for generating a mining path optimal solution set by adopting a hybrid taboo genetic optimization algorithm and evaluating the mining efficiency and the safety risk; the resource collaborative scheduling module is used for establishing a multi-objective optimization model of the equipment-personnel-transportation system by using a space-time constraint particle swarm optimization algorithm; and the visual decision support module supports real-time simulation and adjustment of the mining scheme through a WebGL three-dimensional visual engine and an AR interactive interface. According to the intelligent mining method, accurate calculation of the mining quantity and intelligent mining are achieved, the mining efficiency is improved, safety is guaranteed, resource allocation is optimized, and economic benefits are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of accurate calculation of mineral production volume, and in particular to an intelligent accurate calculation system for mineral production volume based on a three-dimensional geological model. Background Art

[0002] In the field of mineral mining, accurate calculation of mining volume and intelligent mining are crucial for efficient resource utilization, safe production and economic benefits. However, traditional mining volume calculation and related technologies have many limitations and cannot meet the needs of industry development, which are reflected in the following aspects:

[0003] Data collection and processing: In the past, data collection methods were relatively simple, making it difficult to fully obtain complex multi-source heterogeneous data such as geological structure, mineral distribution, equipment working conditions, and personnel positioning. For example, relying only on a small amount of drilling data to infer geological conditions, it is impossible to accurately grasp the details of rock layer changes and mineral distribution, resulting in a deviation in the understanding of the ore body. Moreover, the collected data often lacks effective pre-processing, and has a lot of noise and outliers, which affects the accuracy of subsequent analysis and calculations, resulting in a large error between the geological model constructed based on these data and the actual situation.

[0004] In terms of geological modeling and reserve calculation: Traditional geological modeling methods make it difficult to construct a digital model that can accurately reflect the rock layer properties, fault distribution, and three-dimensional gradient changes in mineral deposits. This results in low accuracy when calculating recoverable reserves, and is unable to provide accurate reserve data and its confidence interval, which brings great uncertainty to mining planning. For example, simple interpolation algorithms cannot fully consider the spatial correlation and uncertainty of geological data, resulting in a large deviation between the reserve calculation results and the actual reserves, which may cause over-exploitation or under-exploitation of resources, affecting the rational development and utilization of mineral resources.

[0005] Mining planning and scheduling: Traditional mining plans are often made based on experience, without fully considering the combined impact of multiple factors such as geological conditions, mining efficiency and safety risks, resulting in unreasonable mining paths, which not only reduces mining efficiency but also increases safety hazards. In terms of resource scheduling, there is a lack of effective coordination mechanisms, and the coordination between equipment, personnel and transportation systems is not close, which can easily lead to problems such as idle equipment, waiting personnel and transportation congestion, resulting in waste of resources and increased costs.

[0006] Therefore, in order to solve the above problems, an intelligent mining volume precision calculation system based on three-dimensional geological model is proposed. Summary of the invention

[0007] The purpose of the present invention is to provide an intelligent mining volume accurate calculation system based on a three-dimensional geological model to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] An intelligent mining volume precision calculation system based on a three-dimensional geological model, comprising:

[0010] Multi-source heterogeneous data acquisition module: It consists of a geological radar array, a laser scanner group, a drilling sensor network and a mining area production equipment monitoring terminal, which enables real-time acquisition and preprocessing of geological structure data, mineral distribution data, equipment operating data and personnel positioning data;

[0011] Dynamic 3D geological modeling module: Adopts the adaptive Kriging interpolation algorithm DAK improved by deep attention mechanism and deep neural network fusion modeling technology to build a digital geological model including rock layer attributes, fault distribution, and three-dimensional gradient changes of mineral levels;

[0012] Accurate reserve calculation module: Integrates the Bayesian-LSTM coupled prediction model and the multi-scale spatial interpolation algorithm to calculate recoverable reserves and their confidence intervals;

[0013] Intelligent mining planning module: Apply the hybrid taboo genetic optimization algorithm HTS-GA to generate the optimal solution set of mining paths and output the comprehensive evaluation index of mining efficiency and safety risk;

[0014] Resource collaborative scheduling module: Use the spatiotemporal constrained particle swarm optimization algorithm STC-PSO to establish a multi-objective optimization model for the equipment-personnel-transportation system;

[0015] Visualization decision support module: deploys WebGL three-dimensional visualization engine and AR interactive interface to support real-time simulation and adjustment of mining plans.

[0016] As a preferred solution, the multi-source heterogeneous data acquisition module includes:

[0017] Distributed fiber optic sensor arrays are deployed along the mine tunnels and integrate temperature, strain, and vibration multi-parameter sensing functions;

[0018] The drone-mounted hyperspectral imaging unit is equipped with a mineral characteristic spectrum database to collect surface data in open-pit mines;

[0019] The data preprocessing unit uses wavelet denoising and outlier robust detection algorithms to clean the raw data.

[0020] As a preferred solution, the dynamic 3D geological modeling module includes:

[0021] Real-time data fusion unit, which seamlessly integrates new exploration data with historical models through spatiotemporal registration algorithms;

[0022] Model self-correction unit, which uses the Bayesian update framework to dynamically correct model parameters;

[0023] The geological uncertainty visualization unit uses transparency gradient mapping technology to present the confidence of reserve distribution.

[0024] As a preferred solution, the precise reserve calculation module includes:

[0025] The variable-scale resource assessment unit achieves accuracy grading of reserve calculations by setting different grid resolutions;

[0026] Time series prediction unit, integrating LSTM neural network to predict the dynamic changes of mineral grade as the mining process progresses;

[0027] The economic recoverability analysis unit calculates economically recoverable reserves by combining the market price fluctuation model with mining cost parameters.

[0028] As a preferred solution, the improved adaptive Kriging interpolation algorithm formula of the dynamic 3D geological modeling module is: ,in, For spatial location Predicted values ​​of geological attributes at The base weights calculated for the traditional Kriging algorithm; and are all attention coefficients based on spatial distance. The target position is measured With Known sample points The degree of correlation between The target position is measured With Known sample points the degree of association between them; For known sample points The geological attribute value at is the number of known sample points; It is the deep feature extracted by the convolutional neural network, indicating that the convolutional neural network is targeting the target spatial position. The extracted A deep feature, From 1 to Get value; is the feature fusion weight coefficient; The number of features extracted for the convolutional neural network.

[0029] As a preferred solution, the Bayesian-LSTM coupling model expression of the precise reserve calculation module is: , in, For the known 1 to Time data Under the conditions of Prediction of mineral grade for a period of time The probability distribution of For the future Predicted value of mineral grade for a period of time; for The geological feature vector at the moment; is the LSTM network parameter set; is the Bayesian hyperparameter; Indicates LSTM network with parameters to process; It is the probability of time series prediction based on LSTM network; is the LSTM network parameter Based on Bayesian hyperparameters The prior distribution of According to 1 to Time data Bayesian hyperparameters The updated posterior distribution; represents the small changes in these two variables during the integration process.

[0030] As a preferred solution, the hybrid optimization algorithm of the intelligent mining planning module includes a two-stage iterative mechanism:

[0031] Phase 1: ;

[0032] Phase 2: ;

[0033] in, is the set of all possible mining paths; For specific mining paths; For path expected production volume; For path The risk factor of is the risk penalty factor; is a collection of resource consumption plans; For specific resource consumption plans; is the number of resource types; For the Class resource consumption; For the The maximum allowed consumption of class resources; is the path continuity constraint weight; is the optimal path obtained in stage 1; is the path obtained in the previous iteration; Represents the distance metric between two paths.

[0034] As a preferred solution, it also includes:

[0035] The blockchain evidence storage unit performs distributed storage of data and operation records of key decision nodes;

[0036] The federated learning driving framework is implemented through the formula ,in, For the Global model parameters for each iteration; The number of mining areas participating in federated learning; For the The mining area is Local model parameters for each iteration; is the learning rate adjustment coefficient; is a sign function, used to take the sign of the value in brackets;

[0037] Digital twin engine, updating equations through physical entity layer pose Layer dynamics equations with virtual models Realize virtual and real synchronous control, where for The predicted position of the physical entity at the moment; for The position of the physical entity at that moment; is the movement speed of the physical entity; is the time interval; is the angle of movement direction; is the position noise; is the mass matrix; is the generalized acceleration vector; is the Coriolis force and centrifugal force matrix; is the gravity vector; To control the input torque; is the external disturbance torque; is the generalized coordinate vector, is the generalized velocity vector.

[0038] As a preferred solution, the spatiotemporal constraint particle swarm algorithm of the resource collaborative scheduling module adopts a dynamic inertia factor: ,in, For particles exist The speed of the moment; For particles exist The speed of the moment; is the current iteration number; is the maximum number of iterations; and is the learning factor, which is a constant; and is a random number uniformly distributed in the interval [0, 1]; For particles The best position found in the past search; For particles exist The location at the moment; The global optimal position searched for all particles; is the constraint gradient influence factor; is the space-time constraint function, , is the number of time constraints; For the The time difference of the time constraints; is the maximum allowed time difference; is the number of resource location constraints; For the The current location of the resource; For the secure location of resources; is the safety radius; is the space-time constraint function Particle Partial derivative with respect to position.

[0039] As a preferred solution, it also includes a dynamic evaluation model for the economic benefits of mining solutions: , is the net present value, which is used to evaluate the economic benefits of mining options; is the total number of time periods during the evaluation period; For the The market price of mineral products in the future; For the The amount of mineral products mined during the period; For the operating costs during the period; is the discount rate; is the initial investment cost; is the number of real options; For the The value of a real option, , To expand the expected value of the project, For additional investment costs; For the The timing of a real option.

[0040] It can be seen from the technical solutions provided by the present invention that the intelligent mining volume precision calculation system based on the three-dimensional geological model provided by the present invention has the following beneficial effects:

[0041] Accurately calculate mining volume: The multi-source heterogeneous data acquisition module comprehensively collects multiple types of data such as geological structure and mineral distribution, and combines with the dynamic three-dimensional geological modeling module to build a high-precision geological model, providing a reliable data basis for the precise reserve calculation module; the precise reserve calculation module integrates advanced models and algorithms, fully considers the temporal and spatial changes and uncertainties of geological data, accurately calculates the recoverable reserves and their confidence intervals, and provides an accurate basis for mineral resource planning and mining, avoiding resource waste or insufficient mining;

[0042] Optimize mining planning and scheduling: The intelligent mining planning module uses a hybrid taboo genetic optimization algorithm to comprehensively consider factors such as expected mining volume, risk factor, resource consumption and path continuity, generate the optimal solution set of the mining path and output comprehensive evaluation indicators to improve mining efficiency and ensure safety; the resource collaborative scheduling module uses a spatiotemporal constrained particle swarm optimization algorithm to establish a multi-objective optimization model of equipment-personnel-transportation system, improve mining coordination, ensure operation safety, improve resource utilization efficiency and reduce mining costs;

[0043] Support visual decision-making: The visual decision support module deploys the WebGL 3D visualization engine and AR interactive interface to simulate and adjust the mining plan in real time; decision-makers can intuitively understand the mining operation, discover and solve problems in a timely manner, improve decision-making efficiency and accuracy, and promote multi-department collaboration;

[0044] Ensure data security and system reliability: The blockchain evidence storage unit performs distributed storage of data and operation records of key decision-making nodes, using blockchain technology to ensure data security, traceability and non-tamperability, facilitate auditing and responsibility tracing, and enhance the basis of cooperative trust; the federated learning driven framework realizes multi-mining area data collaboration and model sharing under the premise of protecting data privacy, and improves the model's generalization ability and adaptability;

[0045] Realize virtual-real synchronization and predictive optimization: The digital twin engine realizes virtual-real synchronization control through equations of the physical entity layer and the virtual model layer, maps the physical entity state in real time, predicts the future state and simulates the mining plan, provides reference for mining decision-making, ensures mining safety, optimizes mining plans and improves mining efficiency;

[0046] Evaluate the economic benefits of mining plans: The dynamic evaluation model for the economic benefits of mining plans comprehensively considers factors such as market price, mining volume, cost, discount rate and real options to quantitatively evaluate the economic benefits of mining plans. It can simulate the changes in economic benefits under different conditions, analyze parameter sensitivity, help decision makers choose the best plan, adjust mining strategies, and improve project profitability and risk resistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The figure is a schematic diagram of the overall structure of an intelligent mining volume precision calculation system based on a three-dimensional geological model according to the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0050] like Figure 1 As shown, an embodiment of the present invention provides an intelligent mining volume precision calculation system based on a three-dimensional geological model, including a multi-source heterogeneous data acquisition module, a dynamic three-dimensional geological modeling module, a precise reserve calculation module, an intelligent mining planning module, a resource collaborative scheduling module and a visual decision support module.

[0051] In this embodiment, the multi-source heterogeneous data acquisition module is composed of a geological radar array, a laser scanner group, a drilling sensor network and a mining area production equipment monitoring terminal, which realizes the real-time acquisition and preprocessing of geological structure data, mineral distribution data, equipment condition data and personnel positioning data;

[0052] Furthermore, the multi-source heterogeneous data acquisition module plays a key role in the intelligent mining volume precision calculation system based on the three-dimensional geological model. It uses a variety of equipment and technologies to collect and pre-process multiple types of data, providing an accurate and comprehensive data basis for subsequent modules, which is of great significance for the realization of accurate calculation of mining volume and intelligent mining. Specifically:

[0053] 1. Hardware composition and data collection scope:

[0054] Geological radar array: This array is rationally laid out in the mining area. It transmits and receives electromagnetic waves, penetrates the strata, and obtains underground geological structure information. It can effectively detect the distribution of rock strata, the location and direction of faults, etc., and provide key geological structure data for the subsequent construction of geological models.

[0055] Laser scanner group: Use laser scanning technology to scan the mining area with high precision; it can accurately measure the spatial distribution of minerals, obtain data such as the boundaries, shapes and spatial relationships between different minerals, and provide an accurate basis for analyzing the distribution of minerals;

[0056] Borehole sensor network: The borehole sensor network is deployed in each borehole in the mining area, which can monitor various physical parameters of the rock around the borehole in real time; such as rock hardness, humidity, stress, etc. This data helps to gain a deeper understanding of the geological conditions and is crucial for accurately calculating the amount of mining;

[0057] Mine production equipment monitoring terminal: installed on various production equipment, real-time collection of equipment operating parameters, including equipment working status, operating speed, energy consumption and other equipment operating data; at the same time, it also has personnel positioning function, which can track the location information of personnel in the mine in real time, ensure personnel safety and facilitate production scheduling;

[0058] 2. Expand equipment and functions:

[0059] Distributed fiber optic sensor array: It is laid along the mine tunnels and integrates the multi-parameter sensing functions of temperature, strain, and vibration. It can monitor the temperature changes around the tunnels in real time to determine whether there are abnormal heat sources, such as possible fire hazards. By monitoring the strain, it can promptly discover the deformation trend of the tunnel surrounding rock and prevent tunnel collapse. The perception of vibration can monitor the vibration generated by nearby mining activities or geological activities to provide protection for safe production.

[0060] UAV-mounted hyperspectral imaging unit: equipped with a mineral characteristic spectrum database, the drone is used to fly over the open-pit mine to collect data; based on the unique spectral characteristics of different minerals, the surface data of the open-pit mine is collected by comparing with the database, and the mineral type, distribution range, content and other information are accurately identified;

[0061] 3. Data preprocessing capabilities:

[0062] Data cleaning algorithm: The data preprocessing unit uses a wavelet noise reduction algorithm to effectively remove noise signals mixed in during data collection and improve the quality and accuracy of data. At the same time, it uses an outlier robust detection algorithm to accurately identify and process outliers in the data, ensuring that the collected data is reliable and stable, and avoiding interference of abnormal data on subsequent calculations and analysis.

[0063] Data transmission and integration: After pre-processing, the collected data are integrated into geological structure data, mineral distribution data, equipment operating data, personnel location data, etc. through a stable data transmission channel, providing a unified and standardized data foundation for subsequent dynamic three-dimensional geological modeling, precise reserve calculation and other modules;

[0064] 4. Application scenario advantages:

[0065] Ensure accurate construction of geological models: By collecting rich geological structure and mineral distribution data, detailed and accurate raw data is provided for the dynamic 3D geological modeling module; the constructed digital geological model can more realistically reflect the underground geological conditions, including rock layer properties, fault distribution, and 3D gradient changes of mineral levels, laying a solid foundation for subsequent accurate calculation of reserves and planning of mining routes;

[0066] Assist in accurate calculation and decision-making of mining volume: The collection of equipment operating data and personnel location data, combined with geological data, plays an important role in the accurate reserve calculation module and intelligent mining planning module; it helps to more accurately calculate the recoverable reserves and their confidence intervals, while generating reasonable mining paths, comprehensively evaluating mining efficiency and safety risks, and providing strong support for intelligent mining decisions;

[0067] Improve the efficiency of resource collaborative scheduling: The collection and integration of multi-source data provides comprehensive information for the resource collaborative scheduling module; by analyzing the real-time status of equipment, personnel and transportation systems, a multi-objective optimization model is established using the spatiotemporal constrained particle swarm optimization algorithm to achieve reasonable allocation and efficient scheduling of resources, thereby improving the coordination and efficiency of the entire mining operation.

[0068] In this embodiment, the dynamic three-dimensional geological modeling module uses the adaptive Kriging interpolation algorithm DAK improved by the deep attention mechanism and the deep neural network fusion modeling technology to construct a digital geological model including the three-dimensional gradient changes of rock layer attributes, fault distribution, and mineral grade;

[0069] Furthermore, the dynamic 3D geological modeling module is a key component of the intelligent mining volume precision calculation system based on the 3D geological model. It uses advanced algorithms and technologies to process and analyze geological data, and builds a digital model that can reflect the geological conditions in real time, providing an important model basis for subsequent precise reserve calculations and mining planning, which is specifically reflected in:

[0070] 1. Modeling technology and algorithm:

[0071] The improved adaptive kriging interpolation algorithm (DAK) with deep attention mechanism and deep neural network fusion: The DAK algorithm and deep neural network fusion modeling technology are adopted, combining the advantages of traditional kriging algorithm and convolutional neural network; the basic weights calculated by the traditional kriging algorithm provide a basic framework for subsequent calculations, and the attention coefficient based on spatial distance measures the degree of correlation between the target position and the known sample points, which enhances the attention to key sample points; the deep features extracted by the convolutional neural network supplement the geological information from a more complex dimension. The combination of the two comprehensively and accurately constructs a digital geological model including rock layer attributes, fault distribution, and three-dimensional gradient changes of mineral grade; the formula of the improved adaptive kriging interpolation algorithm is: ,in, For spatial location Predicted values ​​of geological attributes at The base weights calculated for the traditional Kriging algorithm; and are all attention coefficients based on spatial distance. The target position is measured With Known sample points The degree of correlation between The target position is measured With Known sample points the degree of association between them; For known sample points The geological attribute value at is the number of known sample points; It is the deep feature extracted by the convolutional neural network, indicating that the convolutional neural network is targeting the target spatial position. The extracted A deep feature, From 1 to Get value; is the feature fusion weight coefficient; The number of features extracted for the convolutional neural network;

[0072] 2. Module composition and function:

[0073] Real-time data fusion unit: Use the spatiotemporal registration algorithm to seamlessly integrate the newly acquired exploration data with the historical model; during the continuous exploration of the mining area, the newly acquired data can be quickly and accurately integrated into the existing geological model to ensure that the model can reflect the latest geological information in a timely manner and maintain the timeliness and accuracy of the model;

[0074] Model self-correction unit: Use the Bayesian update framework to dynamically correct model parameters; as new data is continuously added, model parameters will be updated according to the Bayesian principle, so that the model can more accurately fit the actual geological conditions and improve the reliability and adaptability of the model;

[0075] Geological uncertainty visualization unit: uses transparency gradient mapping technology to present the confidence of reserve distribution; through this visualization method, users can intuitively understand the degree of uncertainty of reserve distribution in different regions, providing more intuitive and valuable information for mining decisions;

[0076] 3. Application scenario advantages:

[0077] Providing a reliable model for accurate reserve calculation: The high-precision digital geological model constructed contains rich geological information, providing accurate basic data for the accurate reserve calculation module; integrating the Bayesian-LSTM coupling prediction model and the multi-scale spatial interpolation algorithm, and calculating the recoverable reserves and their confidence intervals based on this model, thus improving the accuracy of reserve calculation;

[0078] Assisted intelligent mining planning: Provide detailed geological model support for the intelligent mining planning module; through the model, you can clearly see the rock layer properties, fault distribution and other information. When applying the hybrid taboo genetic optimization algorithm (HTS-GA) to generate the optimal solution set of the mining path, it can fully consider the geological conditions, output more reasonable mining efficiency-safety risk comprehensive evaluation indicators, and optimize the mining path planning;

[0079] Support visual decision-making: The visual decision support module is based on the dynamic three-dimensional geological model and deploys the WebGL three-dimensional visualization engine and AR interactive interface. Users can view the geological model in real time through these interfaces, simulate and adjust the mining plan in real time, and improve the scientificity and accuracy of decision-making.

[0080] In this embodiment, the precise reserve calculation module integrates the Bayesian-LSTM coupled prediction model and the multi-scale spatial interpolation algorithm to calculate the recoverable reserves and their confidence intervals;

[0081] Furthermore, the precise reserve calculation module is one of the core modules of the intelligent mining volume precise calculation system based on the three-dimensional geological model. It integrates multiple algorithms and models to conduct in-depth analysis and processing of geological data to achieve accurate calculation of recoverable reserves and their confidence intervals, providing key basis for the rational planning and mining of mineral resources. Specifically:

[0082] 1. Core algorithms and models:

[0083] Bayesian-LSTM coupled prediction model: By integrating the Bayesian method and the long short-term memory network (LSTM), the advantages of both are fully utilized; the LSTM neural network has a strong processing ability for time series data and can learn the dynamic change law of mineral grade with the mining process; the Bayesian method is used to deal with the uncertainty of model parameters, and the model is made more accurate and reliable by updating the prior distribution and posterior distribution; its expression is: , in, For the known 1 to Time data Under the conditions of Prediction of mineral grade for a period of time The probability distribution of For the future Predicted value of mineral grade for a period of time; for The geological feature vector at the moment; is the LSTM network parameter set; is the Bayesian hyperparameter; Indicates LSTM network with parameters to process; It is the probability of time series prediction based on LSTM network; is the LSTM network parameter Based on Bayesian hyperparameters The prior distribution of According to 1 to Time data Bayesian hyperparameters The updated posterior distribution; represents the small changes in these two variables during the integration process;

[0084] Multi-scale spatial interpolation algorithm: Combine spatial information of different scales to interpolate mineral grade and reserves; By analyzing and processing geological data at different resolutions, it can more accurately reflect the spatial changes of geological attributes and improve the accuracy of reserve calculation; When calculating reserves, according to the distribution characteristics of geological data, select the appropriate scale for interpolation to make the calculation results more in line with the actual situation;

[0085] 2. Module composition and function:

[0086] Variable-scale resource assessment unit: The accuracy of reserve calculation is graded by setting different grid resolutions. In areas with complex geological conditions and large data changes, high-resolution grids are used for calculation to obtain more accurate reserve data. In areas with relatively stable geological conditions and small data changes, lower-resolution grids are used to improve calculation efficiency. This variable-scale assessment method ensures calculation accuracy while taking into account calculation efficiency.

[0087] Time series prediction unit: The integrated LSTM neural network predicts the dynamic changes of mineral grade as the mining process progresses; as the mining activities proceed, the mineral grade will change due to the influence of various factors; the time series prediction unit uses the LSTM network's ability to learn historical data, combined with real-time geological data and mining information, to predict the future trend of mineral grade changes and provide dynamic input data for reserve calculation;

[0088] Economic recoverability analysis unit: Calculate economic recoverable reserves by combining the market price fluctuation model with mining cost parameters; considering that the mining of mineral resources depends not only on the amount of reserves, but also on the market price and mining cost; this unit obtains market price data in real time, combines mining cost parameters, calculates economic recoverable reserves under different market conditions, and provides an economic reference for mining decisions;

[0089] 3. Application scenario advantages:

[0090] Accurately calculate recoverable reserves: The Bayesian-LSTM coupled prediction model and multi-scale spatial interpolation algorithm are used in combination to fully consider the spatiotemporal changes and uncertainties of geological data, and can more accurately calculate recoverable reserves and their confidence intervals; provide accurate data support for the planning and mining of mineral resources, and avoid resource waste or insufficient mining due to inaccurate reserve calculations;

[0091] Support dynamic mining decisions: The prediction of the dynamic changes of mineral grade by the time series prediction unit and the calculation of economically recoverable reserves by the economic recoverability analysis unit provide strong support for dynamic decision-making in the mining process; mining companies can adjust mining plans and strategies in a timely manner based on this information to improve mining efficiency and economic benefits;

[0092] Adapt to complex geological conditions and market changes: When faced with complex geological conditions and a changing market environment, the precise reserve calculation module can effectively respond to various uncertainties through variable-scale resource assessment, dynamic grade prediction, and economic feasibility analysis; providing reliable technical support for the sustainable development and utilization of mineral resources.

[0093] In this embodiment, the intelligent mining planning module applies the hybrid taboo genetic optimization algorithm HTS-GA to generate the optimal solution set of the mining path and output the mining efficiency-safety risk comprehensive evaluation index;

[0094] Furthermore, the intelligent mining planning module is a key part of the intelligent mining volume precision calculation system based on the three-dimensional geological model. It uses advanced algorithms and models, comprehensively considers multiple factors, and generates scientific and reasonable mining path planning solutions, providing important support for efficient and safe mineral mining, which is specifically reflected in:

[0095] 1. Core algorithms and models:

[0096] Hybrid Taboo Genetic Optimization Algorithm (HTS-GA): This algorithm combines the advantages of taboo search and genetic algorithm. Genetic algorithm simulates natural selection and genetic mechanism to conduct global search for mining paths, and can quickly find the best solution among many possible paths. Taboo search uses taboo table to avoid the algorithm from repeatedly searching the area that has been explored, thereby improving search efficiency and avoiding falling into the local optimal solution. The combination of the two enables the algorithm to search the global space extensively in the process of mining path optimization, and effectively jump out of the local optimality to find a better mining path.

[0097] Two-stage iteration mechanism: It consists of two stages. The goal of stage one is ,in, is the set of all possible mining paths; For specific mining paths; For path expected production volume; For path The risk factor of is the risk penalty factor; this stage is to balance the expected mining volume and risk, and pursue a path that maximizes the expected mining volume and keeps the risk under control; the goal of stage 2 is ,in, is a collection of resource consumption plans; For specific resource consumption plans; is the number of resource types; For the Class resource consumption; For the The maximum allowed consumption of class resources; is the path continuity constraint weight; is the optimal path obtained in stage 1; is the path obtained in the previous iteration; Represents the distance measure between two paths; this stage optimizes resource consumption and path continuity to ensure that resources are used reasonably and the mining path is continuous and stable;

[0098] 2. Module functions:

[0099] Generate the optimal solution set of mining paths: Using the hybrid taboo genetic optimization algorithm and the two-stage iteration mechanism, all possible mining paths are searched and optimized based on multiple factors such as geological conditions, mining volume, risk, resource consumption and path continuity to generate a set of optimal mining path solutions; these paths are the better choices obtained after comprehensively weighing various factors, providing a variety of feasible solutions for actual mining;

[0100] Output mining efficiency-safety risk comprehensive evaluation index: For the generated mining paths, calculate and output the mining efficiency-safety risk comprehensive evaluation index; this index comprehensively reflects the mining efficiency and potential safety risks of each mining path, allowing decision makers to intuitively understand the advantages and disadvantages of different paths, and thus select the most appropriate mining path according to actual needs;

[0101] 3. Application scenario advantages:

[0102] Improve mining efficiency: By optimizing the mining path, reducing unnecessary mining operations and transportation distances, the efficiency of mineral mining is improved; reasonable path planning can make mining equipment run more efficiently, shorten the mining cycle, reduce mining costs, and improve the mining efficiency of mineral resources;

[0103] Ensure mining safety: Risk factors are fully considered when planning mining routes. Risk penalty factors are used to constrain paths with higher risks, avoiding paths that may pose greater safety hazards. This helps reduce safety risks during mining, ensure the safety of workers and the stable production of mines.

[0104] Optimize resource allocation: The resource consumption optimization part of the two-stage iterative mechanism can reasonably plan the use of resources according to the type of resources and the maximum allowable consumption; avoid excessive consumption and waste of resources, achieve efficient allocation of resources, improve resource utilization, and promote the sustainable development of mineral resources.

[0105] In this embodiment, the resource collaborative scheduling module uses the spatiotemporal constrained particle swarm optimization algorithm STC-PSO to establish a multi-objective optimization model for the equipment-personnel-transportation system;

[0106] Furthermore, the resource coordination and scheduling module is an important part of the intelligent mining volume precision calculation system based on the three-dimensional geological model. It uses the spatiotemporal constrained particle swarm optimization algorithm (STC-PSO) to build a multi-objective optimization model of equipment, personnel, and transportation systems, and reasonably allocates various resources in the mining process to improve the overall efficiency and safety of mining operations. Specifically, it is reflected in:

[0107] 1. Core Algorithm:

[0108] Spatiotemporal Constrained Particle Swarm Optimization (STC-PSO): A dynamic inertia factor is used to optimize the search behavior of particles. The formula is: ,in, For particles exist The speed of the moment; For particles exist The speed of the moment; is the current iteration number; is the maximum number of iterations; and is the learning factor, which is a constant; and is a random number uniformly distributed in the interval [0,1]; For particles The best position found in the past search; For particles exist The location at the moment; The global optimal position searched for all particles; is the constraint gradient influence factor;

[0109] Space-time constraint function: , is the number of time constraints; For the The time difference of the time constraints; is the maximum allowed time difference; is the number of resource location constraints; For the The current location of the resource; For the secure location of resources; is the safety radius; is the space-time constraint function Particle The partial derivative of the position; this function comprehensively considers the time constraints and resource location constraints to ensure that resource scheduling meets the time requirements and safety regulations of mining operations;

[0110] 2. Module functions:

[0111] Establish a multi-objective optimization model: Taking equipment, personnel, and transportation systems as optimization objects, comprehensively considering time, space, safety and other constraints, and using the spatiotemporal constraint particle swarm optimization algorithm to establish a multi-objective optimization model; through this model, resources are arranged in a coordinated manner to achieve reasonable allocation and coordinated operation of resources;

[0112] Resource scheduling optimization: Based on the multi-objective optimization model, optimize the use of equipment, deployment of personnel and planning of transportation routes; determine the optimal resource scheduling plan according to the real-time status of mining tasks and resources, improve resource utilization efficiency, reduce waiting time and transportation costs, and ensure the smooth progress of mining operations;

[0113] 3. Application scenario advantages:

[0114] Improve mining coordination: Through the coordinated dispatch of equipment, personnel and transportation systems, each link can work closely together to avoid problems such as idle equipment, redundant personnel or transportation congestion, thereby improving the overall coordination and smoothness of mining operations;

[0115] Ensure operation safety: The resource location constraints and safety radius settings in the spatiotemporal constraint function ensure that resource allocation is always within a safe range; during the scheduling process, the safety distance between equipment and personnel is fully considered to reduce the probability of safety accidents and ensure the safety of operators and the normal operation of equipment;

[0116] Improve resource utilization efficiency: The application of dynamic inertia factors and optimization algorithms enables resource scheduling to be adjusted in real time according to the mining progress and actual conditions; reasonably arrange the work tasks of equipment and personnel, optimize transportation routes, reduce resource waste and consumption, improve resource utilization efficiency, and reduce mining costs.

[0117] In this embodiment, the visual decision support module deploys a WebGL three-dimensional visualization engine and an AR interactive interface to support real-time simulation and adjustment of mining plans;

[0118] Furthermore, the visual decision support module is a key part of the intelligent mining volume precision calculation system based on the three-dimensional geological model to achieve intuitive and efficient decision-making; it uses advanced visualization technology and interactive means to present complex geological data and mining plans to decision makers in an intuitive and easy-to-understand way, assisting them in making scientific decisions, which is of great significance to improving the overall efficiency of mining operations and the accuracy of decision-making. Specifically:

[0119] 1. Core technology:

[0120] WebGL 3D visualization engine: Deploy the WebGL 3D visualization engine, which can efficiently render 3D graphics in web browsers. It uses the parallel computing power of the graphics processing unit (GPU) to present the digital geological model constructed by the dynamic 3D geological modeling module, the mining path generated by the intelligent mining planning module, and the resource distribution data of the resource collaborative scheduling module in a realistic 3D form. Users can access these 3D models anytime and anywhere through the browser to conduct multi-angle and multi-level observations, and clearly understand the geological structure of the mining area, the distribution of ore bodies, and the planning layout of mining operations;

[0121] AR interactive interface: Equipped with AR interactive interface, augmented reality technology combines virtual information with real scenes. At the mine site, workers can use mobile devices or head-mounted display devices, with the help of AR interactive interface, to overlay virtual mining plans, equipment operating parameters, safety warnings and other information on the real mining environment. This allows workers to obtain key information in real time in the actual working environment and interact naturally with virtual models, such as gesture operation and voice commands, to facilitate on-site verification and adjustment of mining plans.

[0122] 2. Module functions:

[0123] Real-time simulation of mining plans: Use the WebGL 3D visualization engine and AR interactive interface to simulate mining plans in real time; simulate the running track of equipment, the mining sequence of ore, transportation routes, etc. during the mining process, and display the entire mining operation process; through simulation, decision makers can discover possible problems in the mining plan in advance, such as equipment collision, unreasonable mining sequence, transportation route congestion, etc., and make timely adjustments and optimizations to avoid these problems in the actual mining process, reducing economic losses and safety risks;

[0124] Plan adjustment and decision support: Support decision makers to make real-time adjustments to mining plans on a visual interface; decision makers can modify mining paths, adjust equipment configurations, optimize resource scheduling, etc. directly on a 3D model or AR interactive interface based on simulation results, actual on-site conditions, and various emergencies; at the same time, the system will provide real-time feedback on the impact of these adjustments on mining efficiency, safety risks, economic benefits, etc., providing comprehensive decision support for decision makers and helping them make more scientific and reasonable decisions;

[0125] 3. Application scenario advantages:

[0126] Improve decision-making efficiency and accuracy: Visual presentation and convenient interactive means enable decision makers to quickly understand complex geological data and mining plans, and intuitively evaluate the pros and cons of different plans; through real-time simulation and adjustment functions, decision makers can make more accurate decisions in a short time, avoiding decision-making errors and delays caused by difficulties in understanding data and non-intuitive plan evaluation in traditional decision-making methods;

[0127] Promote multi-department collaboration: This module provides a unified visualization platform for multiple departments such as geological exploration, mining design, production scheduling, and safety management. Personnel from various departments can share information and work together on this platform to jointly participate in the formulation and optimization of mining plans. For example, geologists can introduce the characteristics of geological structures in detail to mining designers on a visualization interface. Mining designers can adjust mining plans directly on the platform based on this information. Safety managers can check safety hazards in the plans in real time and put forward improvement suggestions, thereby improving the coordination and efficiency of the entire mining operation.

[0128] Enhanced on-site work guidance: At the mine site, the AR interactive interface provides real-time work guidance for workers; workers can obtain virtual operation guides, equipment status information, etc. in real scenes through mobile devices or head-mounted display devices, helping them to complete various work tasks more accurately and efficiently; at the same time, the AR interactive interface can also display safety warning information in real time, reminding workers to pay attention to safety and reducing the probability of safety accidents.

[0129] An intelligent mining volume precision calculation system based on a three-dimensional geological model, which also includes a blockchain evidence storage unit, a federated learning drive framework, a digital twin engine, and a dynamic evaluation model for the economic benefits of mining plans;

[0130] Furthermore, the blockchain evidence storage unit performs distributed storage of data and operation records of key decision nodes, specifically:

[0131] The blockchain evidence storage unit is an important component of the system of the present invention. It uses the characteristics of blockchain technology to perform distributed storage of data and operation records of key decision-making nodes, providing data security, traceability and non-tamperability for the entire intelligent mining volume precision calculation system; including:

[0132] 1. Core technology and principle:

[0133] Blockchain is a decentralized distributed ledger technology that uses cryptography to ensure the security of data transmission and access. Based on this technology, the blockchain evidence storage unit packages the data and operation records of key decision nodes into "blocks". Each block contains multiple records within a certain period of time and the hash value of the previous block, thus forming a chain structure. This chain structure makes the data extremely coherent and traceable. Any tampering with the data will cause the hash value of subsequent blocks to change, which can be easily discovered.

[0134] 2. Features:

[0135] Distributed data storage: In the system of the present invention, the blockchain evidence storage unit does not rely on a single server or storage device, but stores data in multiple nodes in a distributed manner; these nodes can be different computer devices and servers in the mining area, and can even be extended to the entire network range of the participating system; this distributed storage method greatly improves the security and reliability of data. Even if some nodes fail or are attacked, data can still be obtained from other nodes without causing data loss;

[0136] Operation record storage: Detailed record storage of operation records of key decision nodes in the system; for example, in the process of model update operation in the dynamic 3D geological modeling module, reserve calculation completion in the precise reserve calculation module, and mining path decision generation in the intelligent mining planning module, relevant operation steps, parameter settings, decision basis and other information will be recorded; these records will not only help the subsequent review and analysis of the decision-making process, but also provide an accurate basis for responsibility tracing;

[0137] Data cannot be tampered with: The cryptographic technology of blockchain ensures that the evidence data cannot be tampered with. Once the data is recorded on the blockchain, it is difficult to be illegally modified. Because tampering with data requires modifying the information in multiple blocks at the same time and controlling more than half of the nodes, which is almost impossible in practice. This feature ensures the authenticity and integrity of the data, making the decisions based on this data more credible.

[0138] 3. Application scenario advantages:

[0139] Ensure data security and reliability: In the process of mineral mining, data security is of vital importance; the blockchain evidence storage unit can effectively prevent data from being maliciously tampered with or lost, and ensure the safe storage of geological data, mining decision data, etc.; whether it is dealing with internal data management risks or external network attack threats, it can ensure data reliability and provide stable data support for subsequent mining operations;

[0140] Facilitates auditing and responsibility tracing: In complex mining operations, multiple links and numerous personnel are involved in decision-making and operations. When problems arise or audits are required, blockchain evidence storage units can provide complete and accurate operation records and data change history. By tracing these records, we can clearly understand the process of making each decision and the person responsible, which is convenient for responsibility identification and problem troubleshooting, and improves the management efficiency and standardization of mining operations.

[0141] Enhance the trust foundation for cooperation: In mineral mining projects involving multi-party cooperation, such as sharing geological data, mining plans and other information with partners, the existence of blockchain evidence units can enhance the trust between the parties; because all parties can verify the authenticity and integrity of the data through blockchain technology, there is no need to worry about the data being tampered with or forged, thus providing a more solid trust foundation for cooperation and promoting the smooth progress of cooperation;

[0142] The federated learning driven framework is the key part of the present invention to promote data collaboration and model optimization in multiple mining areas. Through a unique algorithm mechanism, it realizes knowledge sharing while protecting the privacy of data in each mining area, and improves the performance and adaptability of the entire system. The federated learning driven framework uses the formula ,in, For the The global model parameters of the iteration round, which integrate the local model parameter information of each mining area and represent the comprehensive learning results of the entire federated learning system in this iteration round; The number of mining areas participating in federated learning reflects the scope of collaboration; For the The mining area is The local model parameters of each round of iteration reflect the model characteristics of each mining area based on its own data training; The learning rate adjustment coefficient is used to control the step size of global model parameter updates and balance the convergence speed and accuracy of the model; is a sign function, which is used to take the sign of the value in the brackets. Its function is to guide the update direction of the global model parameters according to the difference direction between the local model parameters of each mining area and the global model parameters;

[0143] Framework function: It can realize data collaboration and model sharing among multiple mining areas under the premise of protecting data privacy. Each mining area can use its own geological data, mining data, etc. to train models locally and generate local model parameters. These parameters will not directly expose the original data, but will be integrated with the model parameters of other mining areas through the federated learning algorithm to jointly optimize the global model. By continuously iterating and updating the global model parameters, the model can learn the characteristics and laws of different mining areas, and improve the generalization ability and adaptability of the model. There are differences in geological conditions and mining conditions in different mining areas. The federated learning-driven framework can integrate these diverse information, so that the model can perform better in the face of various complex situations.

[0144] Advantages of application scenarios: For enterprises or institutions with multiple mining areas, the federated learning-driven framework avoids privacy risks and security risks caused by centralized storage and transmission of data; each mining area does not need to upload sensitive data to the central server, reducing the risk of data leakage and meeting the requirements of data security and privacy protection; through data collaboration among multiple mining areas, the global model can integrate the advantageous information of different mining areas, which is more accurate and comprehensive than the model independently trained in a single mining area; in tasks such as calculating mining volume and planning mining routes, it can provide more reliable results and decision support, thereby improving the efficiency and benefits of the entire mining operation;

[0145] The digital twin engine is a key technical module in the present invention to achieve real-time synchronization and coordinated control between the physical world and the virtual world. It provides more intuitive and accurate decision support and process monitoring means for intelligent mining by establishing a close connection between physical entities and virtual models. Specifically:

[0146] Core equations and principles: Update equations with the help of physical entity layer pose Layer dynamics equations with virtual models Realize virtual-real synchronous control at the physical entity layer. for The predicted position of the physical entity at a given moment, based on its position at the previous moment , combined with the movement speed of the physical entity , Movement direction angle and time interval Make updates; is the position noise, simulating the uncertainty factors in the actual environment; at the virtual model layer, the mass matrix , generalized acceleration vector , Coriolis force and centrifugal force matrix , gravity vector , control input torque and external disturbance torque Parameters such as and jointly describe the dynamic characteristics of the virtual model, ensuring that the virtual model can accurately reflect the motion and force of the physical entity; is the generalized coordinate vector, is the generalized velocity vector;

[0147] Engine functions: Real-time mapping of the state of physical entities to virtual models. By continuously collecting information such as the position and speed of physical entities, using the physical entity layer posture update equation, the corresponding parameters are synchronously updated in the virtual model, so that the virtual model and the physical entity are highly consistent in state; based on the dynamic equation of the virtual model, the future state of the physical entity can be predicted and potential problems can be discovered in advance; during the operation of mining equipment, the wear of equipment components and operating stability can be predicted to provide a basis for equipment maintenance and scheduling; different mining plans and working conditions can be simulated in a virtual environment to evaluate their impact on physical entities and the entire mining system; by comparing and analyzing different simulation results, references can be provided for actual mining decisions and mining plans can be optimized;

[0148] Advantages of application scenarios: In the process of mineral mining, by real-time monitoring of the synchronization between physical entities and virtual models, problems such as equipment failures and abnormal operations of personnel can be discovered in a timely manner; the prediction function of the virtual model can warn of potential risks in advance and help staff take corresponding measures to ensure the safety of mining operations; the digital twin engine is used to simulate and optimize mining plans in a virtual environment to avoid the costs and risks of large-scale experiments in actual mining; by comparing different simulation results, the optimal mining plan can be selected to improve mining efficiency and reduce resource waste;

[0149] The dynamic evaluation model of the economic benefits of mining schemes is a key tool used in the present invention to measure the economic feasibility of mining schemes. By comprehensively considering various factors, the economic benefits of mining schemes are quantitatively evaluated, providing an important basis for mining decisions.

[0150] Core formula and parameters of the dynamic evaluation model of economic benefits of mining schemes: The net present value (NPV) formula is used for economic benefit evaluation, and the formula is: , The net present value is used to evaluate the economic benefits of mining plans. A positive value indicates that the plan is economically feasible, while a negative value indicates that there may be losses. The total time period of the assessment period clarifies the time range of the assessment. Considering that mineral mining is a long-term process, a reasonable assessment period can more comprehensively reflect the economic impact of the mining plan; For the The market price of mineral products in the period, For the The amount of mineral products mined in each period. These two parameters directly affect the income of each period; For the The operating costs during the mining period include equipment maintenance, manpower, energy consumption, etc., which reflect the cost expenditure during the mining process; The discount rate is used to convert future cash flows into present values, taking into account the time value of money and reflecting the market's investment return rate and risk level; The initial investment cost covers the one-time investment in equipment purchase and infrastructure construction in the early stage of mining; is the number of real options; For the The value of a real option, , To expand the expected value of the project, For additional investment costs; For the Real options take into account the possibility of future project expansion or adjustment, making the evaluation more realistic.

[0151] Model function: comprehensively evaluate the economic benefits of the mining scheme during the entire evaluation period, comprehensively consider the income, cost, initial investment and real option value, and obtain a quantitative net present value indicator to intuitively reflect the economic feasibility of the scheme; by adjusting the parameters in the model, such as market price, mining volume, cost, discount rate, etc., simulate the changes in economic benefits under different market environments and mining conditions, and provide direction for the optimization of mining schemes; analyze the sensitivity of each parameter to the net present value, determine which factors have a greater impact on economic benefits, and help decision makers focus on and manage key factors to reduce economic risks;

[0152] Advantages of application scenarios: When formulating mining plans, the model is used to evaluate the economic benefits of different plans, and plans with higher net present values ​​are selected to ensure the economic rationality of the mining project and improve the return on investment. During the mining process, as market prices fluctuate and costs change, the model is used to re-evaluate the economic benefits and adjust the mining strategy in a timely manner, such as adjusting the mining volume, optimizing the cost structure, or considering the execution of real options to ensure the profitability of the project. When faced with an uncertain market environment and complex mining conditions, the real options in the model take into account the flexibility and potential value of the project, avoiding short-sighted decisions due to ignoring possible future changes, and improving the ability of mining projects to cope with risks.

[0153] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent mining volume precision calculation system based on a three-dimensional geological model, characterized by: include: Multi-source heterogeneous data acquisition module: It consists of a geological radar array, a laser scanner group, a drilling sensor network and a mining area production equipment monitoring terminal, which realizes the real-time acquisition and preprocessing of geological structure data, mineral distribution data, equipment condition data and personnel positioning data; Dynamic 3D geological modeling module: Adopts the adaptive Kriging interpolation algorithm DAK improved by deep attention mechanism and deep neural network fusion modeling technology to build a digital geological model including rock layer attributes, fault distribution, and three-dimensional gradient changes of mineral levels; Accurate reserve calculation module: Integrates the Bayesian-LSTM coupled prediction model and the multi-scale spatial interpolation algorithm to calculate recoverable reserves and their confidence intervals; Intelligent mining planning module: Apply the hybrid taboo genetic optimization algorithm HTS-GA to generate the optimal solution set of mining paths and output the comprehensive evaluation index of mining efficiency and safety risk; Resource collaborative scheduling module: Use the spatiotemporal constrained particle swarm optimization algorithm STC-PSO to establish a multi-objective optimization model for the equipment-personnel-transportation system; Visualization decision support module: deploys WebGL three-dimensional visualization engine and AR interactive interface to support real-time simulation and adjustment of mining plans.

2. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The multi-source heterogeneous data acquisition module includes: Distributed fiber optic sensor arrays are deployed along the mine tunnels and integrate temperature, strain, and vibration multi-parameter sensing functions; The drone-mounted hyperspectral imaging unit is equipped with a mineral characteristic spectrum database to collect surface data in open-pit mines; The data preprocessing unit uses wavelet denoising and outlier robust detection algorithms to clean the raw data.

3. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The dynamic three-dimensional geological modeling module comprises: Real-time data fusion unit, which seamlessly integrates new exploration data with historical models through spatiotemporal registration algorithms; Model self-correction unit, which uses the Bayesian update framework to dynamically correct model parameters; The geological uncertainty visualization unit uses transparency gradient mapping technology to present the confidence of reserve distribution.

4. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The precise reserve calculation module includes: The variable-scale resource assessment unit achieves accuracy grading of reserve calculations by setting different grid resolutions; Time series prediction unit, integrating LSTM neural network to predict the dynamic changes of mineral grade as the mining process progresses; The economic recoverability analysis unit calculates economically recoverable reserves by combining the market price fluctuation model with mining cost parameters.

5. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The improved adaptive Kriging interpolation algorithm formula of the dynamic 3D geological modeling module is: ,in, For spatial location Predicted values ​​of geological attributes at The base weights calculated for the traditional Kriging algorithm; and are all attention coefficients based on spatial distance. The target position is measured With Known sample points The degree of correlation between The target position is measured With Known sample points the degree of association between them; For known sample points The geological attribute value at is the number of known sample points; It is the deep feature extracted by the convolutional neural network, indicating that the convolutional neural network is targeting the target spatial position. The extracted A deep feature, From 1 to Get value; is the feature fusion weight coefficient; The number of features extracted for the convolutional neural network.

6. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The Bayesian-LSTM coupling model expression of the precise reserve calculation module is: ， in, For the known 1 to Time data Under the conditions of Prediction of mineral grade for a period of time The probability distribution of For the future Predicted value of mineral grade for a period of time; for The geological feature vector at the moment; is the LSTM network parameter set; is the Bayesian hyperparameter; Indicates LSTM network with parameters to process; It is the probability of time series prediction based on LSTM network; is the LSTM network parameter Based on Bayesian hyperparameters The prior distribution of According to 1 to Time data Bayesian hyperparameters The updated posterior distribution; represents the small changes in these two variables during the integration process.

7. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The hybrid optimization algorithm of the intelligent mining planning module includes a two-stage iterative mechanism: Phase 1: ; Phase 2: ; in, is the set of all possible mining paths; For specific mining paths; For path expected production volume; For path The risk factor of is the risk penalty factor; is a collection of resource consumption plans; For specific resource consumption plans; is the number of resource types; For the Class resource consumption; For the The maximum allowed consumption of class resources; is the path continuity constraint weight; is the optimal path obtained in stage 1; is the path obtained in the previous iteration; Represents the distance metric between two paths.

8. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: Also includes: The blockchain evidence storage unit performs distributed storage of data and operation records of key decision nodes; The federated learning driving framework is implemented through the formula ,in, For the Global model parameters for each iteration; The number of mining areas participating in federated learning; For the The mining area is Local model parameters for each iteration; is the learning rate adjustment coefficient; is a sign function, used to take the sign of the value in brackets; Digital twin engine, updating equations through physical entity layer pose Layer dynamics equations with virtual models Realize virtual and real synchronous control, where for The predicted position of the physical entity at the moment; for The position of the physical entity at that moment; is the movement speed of the physical entity; is the time interval; is the movement direction angle; is the position noise; is the mass matrix; is the generalized acceleration vector; is the Coriolis force and centrifugal force matrix; is the gravity vector; To control the input torque; is the external disturbance torque; is the generalized coordinate vector, is the generalized velocity vector.

9. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1 is characterized by: The spatiotemporal constraint particle swarm algorithm of the resource collaborative scheduling module adopts a dynamic inertia factor: ,in, For particles exist The speed of the moment; For particles exist The speed of the moment; is the current iteration number; is the maximum number of iterations; and is the learning factor, which is a constant; and is a random number uniformly distributed in the interval [0, 1]; For particles The best position found in the past search; For particles exist The location at the moment; The global optimal position searched for all particles; is the constraint gradient influence factor; is the space-time constraint function, , is the number of time constraints; For the The time difference of the time constraints; is the maximum allowed time difference; is the number of resource location constraints; For the The current location of the resource; For the secure location of resources; is the safety radius; is the space-time constraint function Particle Partial derivative with respect to position.

10. The intelligent mining volume accurate calculation system based on three-dimensional geological model according to claim 1, characterized in that: It also includes a dynamic evaluation model for the economic benefits of mining plans: , is the net present value, which is used to evaluate the economic benefits of mining options; is the total number of time periods during the evaluation period; For the The market price of mineral products in the future; For the The amount of mineral products mined during the period; For the operating costs during the period; is the discount rate; is the initial investment cost; is the number of real options; For the The value of a real option, , To expand the expected value of the project, For additional investment costs; For the The timing of a real option.

Images