A method for managing the reserves of mine geological resources

Abstract

The present invention relates to the technical field of mine exploitation planning and management, and particularly to a method for managing mine geological resource reserves. The method comprises the following steps: collecting multi-source mine geological data and performing preprocessing on the multi-source data to obtain fused geological data; extracting geological semantic features from the fused geological data and performing semantic reconstruction to obtain semantic geological units; constructing a variogram using the fused geological data; calculating reserve probabilities according to the variogram to obtain unit probability data; performing Bayesian inference based on the unit probability data and generating unit reserve probabilities to obtain unit reserve probabilities; constructing mining units from the semantic geological units to obtain mining units; and constructing an exploitation optimization objective function according to the mining units and the unit reserve probabilities to obtain a comprehensive objective function. The present invention accurately reflects the changes in mine resources through a probabilistic reserve assessment method, improves resource utilization rate, and reduces production costs.

Description

Technical Field

[0001] The present invention relates to the technical field of mine exploitation planning and management, and particularly to a method for managing mine geological resource reserves. Background Art

[0002] The existing methods for managing mine geological resource reserves usually construct static geological models, which cannot reflect the changes in geological conditions during the mine exploitation process and lack a dynamic update mechanism for mine resource reserve management. As the exploitation progresses, new geological information is continuously revealed, but the static model cannot absorb this new information in a timely manner, resulting in a decline in model accuracy and affecting the accuracy of reserve assessment and exploitation planning.

[0003] Most of the existing methods use deterministic models for reserve assessment. For example, in the block model method, the grade of each block is regarded as a definite value. However, there is great uncertainty in geological data itself, such as sparse distribution of borehole data, measurement errors, etc. These uncertainties are not fully considered in traditional methods, affecting the accuracy of resource reserve management.

[0004] In summary, the problems of the existing methods, such as the lack of a dynamic update mechanism and the single reserve assessment method, need to be solved urgently. Summary of the Invention

[0005] Based on this, it is necessary to provide a method for managing mine geological resource reserves to solve at least one of the above technical problems.

[0006] To achieve the above object, a method for managing mine geological resource reserves includes the following steps:

[0007] Step S1: Collect multi-source mine geological data, perform preprocessing on the multi-source data to obtain fused geological data; extract geological semantic features from the fused geological data and perform semantic reconstruction to obtain semantic geological units;

[0008] Step S2: Use the fused geological data to construct a variogram; calculate the reserve probability according to the variogram to obtain unit probability data; perform Bayesian inference based on the unit probability data and generate unit reserve probability to obtain unit reserve probability;

[0009] Step S3: Construct exploitation units for the semantic geological units to obtain exploitation units; construct an exploitation optimization objective function according to the exploitation units and the unit reserve probability to obtain a comprehensive objective function; construct a constraint function according to the exploitation units to obtain a constraint function; perform multi-objective optimization solution on the exploitation units according to the comprehensive objective function and the constraint function, and evaluate the exploitation plan to obtain a target exploitation plan;

[0010] Step S4: Simulate the mining process according to the target mining plan to obtain a mining performance dataset; conduct mining difference analysis on the mining performance dataset to obtain a difference analysis report; and correct the geological model based on the difference analysis report and semantic geological units to obtain a corrected geological model.

[0011] The present invention constructs semantic geological units by integrating multi-source geological data and extracting key geological semantic features. First, obtain geological data from different sources, such as borehole data, geological maps, geophysical data, etc., and perform preprocessing, including data cleaning, format conversion, coordinate unification, etc., to eliminate data differences and form fused geological data. Then, extract geological semantic features from the fused dataset, such as lithology, mineral composition, structure and texture, etc., and use these features to perform semantic reconstruction of the geological space, divide semantic geological units with similar geological attributes, and finally form semantic geological units. Based on the semantic geological units, conduct variogram model analysis to quantify the spatial variability and correlation of geological attributes. Then, use the variogram to construct a reserve probability model, such as a probability model based on a Bayesian network, to express the relationship between geological attributes, spatial correlation, and reserves in a probabilistic form. Finally, use the Bayesian inference method, combined with prior information and the unit probability model, to generate unit reserve probabilities and probabilistically estimate the reserves of each semantic geological unit. According to the semantic geological units and mine mining technical parameters, construct mining units and divide the ore body into mineable units. Then, based on the mining units and unit reserve probabilities, construct a mining optimization objective function, such as maximizing the net present value, minimizing the mining cost, etc., and transform the mining constraint conditions into mathematical expressions to form a constraint function. Finally, use multi-objective optimization algorithms, such as genetic algorithms, simulated annealing algorithms, etc., to perform multi-objective optimization solution on the mining units, and evaluate the obtained mining plan, and finally determine the target mining plan. According to the target mining plan, semantic geological units, and mining units, construct a mine mining simulation environment. Then, conduct mining process simulation in the simulation environment and collect mining performance data, such as production volume, grade, cost, etc. Next, compare and analyze the simulated mining performance data with the actual mining data to find the differences, and determine the parameters that need to be corrected in the geological model according to the difference analysis report. Finally, select a suitable correction method according to the correction target, adjust the model parameters of the semantic geological units, and conduct model verification and evaluation to finally generate a corrected set of geological models. Therefore, the present invention provides a method for managing mine geological resource reserves, which effectively solves the problems of the existing geological resource reserve management method lacking a dynamic update mechanism and a single reserve evaluation method by constructing a dynamically feedback closed-loop system and adopting a probabilistic reserve evaluation method. It accurately reflects the changes in mine resources, improves resource utilization rate, and reduces production costs.

[0012] Preferably, step S1 includes the following steps:

[0013] Step S11: Collect multi-source mine geological data; perform multi-source data preprocessing on the multi-source mine geological data to obtain fused geological data;

[0014] Step S12: Construct an initial 3D geological model for the fused geological data and conduct preliminary grid division to obtain an initial geological grid model;

[0015] Step S13: Extract geological semantic features from the fused geological data according to the initial geological grid model to obtain a geological feature vector set;

[0016] Step S14: Perform geological semantic segmentation on the initial geological grid model according to the geological feature vector set to obtain an initial semantic unit set;

[0017] Step S15: Optimize the initial semantic unit set to obtain semantic geological units.

[0018] By acquiring and preprocessing multi-source mine geological data, the present invention forms fused geological data, eliminates the differences in data sources and formats, provides a unified and standardized data basis for subsequent geological modeling and analysis, and improves the consistency and reliability of the data. Based on the fused geological data, an initial 3D geological model is constructed and grid division is carried out, digitizing the geometric shapes and spatial distributions of ore bodies and surrounding rocks, and forming an initial geological grid model that is convenient for computer processing and analysis, laying a foundation for subsequent geological semantic feature extraction and segmentation. Geological semantic features are extracted from the initial geological grid model, and these features are quantified into a geological feature vector set, realizing the digital expression of geological information and providing necessary input data for subsequent geological semantic segmentation. Using the geological feature vector set to perform geological semantic segmentation on the initial geological grid model, the ore body is divided into different initial semantic units, initially realizing the classification and identification of geological bodies, and providing basic units for subsequent reserve estimation and mining planning. By optimizing the initial semantic unit set, such as merging similar units, splitting complex units, adjusting boundaries, etc., the semantic geological units are made to better conform to geological laws and actual mining conditions, improving the accuracy and practicality of the geological model, and providing a more reliable geological unit division result for subsequent reserve management.

[0019] Preferably, step S2 includes the following steps:

[0020] Step S21: Determine the key geological attributes of the semantic geological units;

[0021] Step S22: Perform reserve probability calculation based on the Bayesian network structure on the semantic geological units according to the variogram to obtain unit probability data;

[0022] Step S23: Perform Bayesian inference based on the unit probability data and update the parameters to obtain updated probability data;

[0023] Step S24: Generate a reserve probability distribution for the semantic geological unit based on the updated probability data to obtain a unit reserve distribution;

[0024] Step S25: Generate a unit reserve probability for the unit reserve distribution to obtain a unit reserve probability.

[0025] The present invention constructs a reserve probability model with a Bayesian network structure based on variograms and semantic geological units, expresses the complex relationship among geological attributes, spatial correlation, and reserves in the form of probability, and provides an effective method for more comprehensive and accurate assessment of reserve uncertainty. By using Bayesian inference and newly added exploration data to update the parameters of the unit probability data, updated probability data is obtained, enabling the reserve probability model to continuously learn and improve, enhancing the prediction accuracy and adaptability of the model, and making the reserve assessment results closer to the actual situation. The reserve probability distribution of each semantic geological unit is generated based on the updated probability data, expressing the reserves in the form of a probability distribution rather than a single deterministic value, more comprehensively reflecting the uncertainty of the reserves, and providing more reliable information for subsequent mining decisions. By sorting and standardizing the unit reserve distribution, a unit reserve probability is generated, storing the reserve uncertainty information of all units in a structured form, providing a convenient data interface and more standardized data input for subsequent mining optimization, and improving the efficiency and accuracy of mining planning.

[0026] Preferably, step S3 includes the following steps:

[0027] Step S31: Obtain the mining technical parameters and mining constraint conditions of the mine; construct mining units for the semantic geological units according to the mining technical parameters of the mine to obtain mining units;

[0028] Step S32: Construct an optimization objective function for mining according to the mining units and the unit reserve probability to obtain a comprehensive objective function;

[0029] Step S33: Convert the mining constraint conditions into mathematical expressions and construct constraint functions according to the mining units and the mining technical parameters of the mine to obtain constraint functions;

[0030] Step S34: Use the comprehensive objective function and the constraint functions as the input of the optimization algorithm, and perform multi-objective optimization solution for the mining units based on the multi-objective optimization algorithm to obtain a solution set of mining sequences;

[0031] Step S35: Generate a mining plan for the solution set of mining sequences to obtain a mining plan;

[0032] Step S36: Evaluate the mining plan and conduct Pareto optimal solution analysis to obtain the target mining plan.

[0033] The present invention obtains the mining technical parameters and constraint conditions of the mine, constructs the mining units for the semantic geological units based on these parameters, and obtains mining units that are more in line with the actual mining situation, laying a foundation for subsequent mining optimization and making the optimization results more practical. The mining optimization objective function is constructed according to the mining units and the probability of unit reserves, integrating multiple mining objectives of the mine (such as maximizing reserves, grade balance, geological structure stability, mining efficiency) into a comprehensive objective function, realizing the synchronous optimization of multiple objectives and making the mining plan more in line with the overall interests of the mine. The mining constraint conditions are transformed into mathematical expressions, and a constraint function is constructed to incorporate various limiting factors in actual mining into the optimization model, ensuring the feasibility of the mining plan and avoiding situations that violate safety regulations or exceed the production capacity of the mine. The multi-objective optimization algorithm is used to optimize and solve the mining units, obtaining the Pareto optimal solution set, providing a series of mining plans that take into account multiple objectives, providing more choices for decision-makers, and making the decision-making process more flexible and scientific. The mining sequence solution set is transformed into a specific mining plan, refining the mining plan, including the mining sequence, time arrangement, resource allocation, etc., making the mining plan more operable and guiding, and facilitating actual implementation. By evaluating the mining plan and conducting Pareto optimal solution analysis, the final target mining plan is obtained, comprehensively considering multiple objectives and constraint conditions, realizing the optimal allocation and sustainable mining of mine resources, and maximizing the economic and social benefits of the mine. Brief Description of the Drawings

[0034] Figure 1 It is a schematic diagram of the step process of a method for managing the reserves of mine geological resources.

[0035] The implementation, functional characteristics and advantages of the object of the present invention will be further described in conjunction with the embodiments with reference to the drawings. Detailed Embodiments

[0036] The technical method of the present invention will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0037] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0038] It should be understood that although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0039] In the embodiments of the present invention, with reference to Figure 1 as shown, it is a schematic diagram of the step flow of the method for managing the reserves of mine geological resources of the present invention. In this example, the method for managing the reserves of mine geological resources includes the following steps:

[0040] Step S1: Collect multi-source data of mine geology, and perform preprocessing of multi-source data to obtain fused geological data; extract geological semantic features from the fused geological data, and perform semantic reconstruction to obtain semantic geological units;

[0041] In the embodiments of the present invention, first, comprehensively collect borehole data (lithology, grade, coordinates, etc.), geophysical data (electrical method data, etc.), remote sensing data, existing geological maps and reports, etc. within the mine area. Perform preprocessing on these multi-source heterogeneous data, including data cleaning (processing missing values and outliers), format conversion (unifying data formats), coordinate system conversion and registration (unifying to the UTM coordinate system). Then, use the grade data in the borehole data and adopt the Kriging interpolation method to generate a three-dimensional grade distribution map to form fused geological data. Next, based on the fused geological data and borehole data, adopt an implicit modeling method (radial basis function interpolation to construct the formation interface, moving surface fitting to construct the fault surface) to construct an initial three-dimensional geological model, and perform regular grid division (for example, 5m * 5m*(A 5m cube unit) to obtain an initial geological grid model. Subsequently, the geometric features (center point coordinates, volume, shape index, distance to interfaces and faults), property features (average grade, main lithology, density), topological features (adjacency matrix), and geological structure features (distance to faults, fault influence degree, electrical method anomaly features) of each grid unit are extracted, and these features are combined into a feature vector. Finally, based on the set of geological feature vectors, the K-means clustering algorithm is used for geological semantic segmentation, and optimization (isolated unit processing, unit merging, splitting, boundary adjustment) is performed according to geological prior knowledge (such as stratigraphic contact relationships) to obtain semantic geological units.

[0042] Step S2: Construct a variogram using the fused geological data; calculate the reserve probability according to the variogram to obtain unit probability data; perform Bayesian inference based on the unit probability data and generate the unit reserve probability to obtain the unit reserve probability;

[0043] In the embodiment of the present invention, the borehole data corresponding to the key geological attributes in the fused geological data (FGD) is used for variogram analysis. Taking the copper grade in a copper deposit as an example: 1. Experimental variogram calculation: Select a semantic geological unit (such as a certain ore body unit), and extract the copper grade data of all boreholes within this unit from the FGD. Calculate the experimental variogram values at different lag distances (h). . The specific calculation formula is: , where N(h) is the number of sample point pairs with a lag distance of h, Z(xi) is the grade value at position xi, and Z(xi + h) is the grade value at a position h away from xi. Calculate the experimental variogram values in multiple directions (such as 0°, 45°, 90°, 135°) and at different lag distances, and draw an experimental variogram graph. 2. Theoretical variogram model fitting: According to the shape of the experimental variogram graph, select a suitable theoretical variogram model for fitting. Commonly used theoretical variogram models include the spherical model, exponential model, Gaussian model, etc. For example, if the experimental variogram graph shows an obvious "sill effect", the spherical model can be selected for fitting. The formula for the spherical model is: when h < a; , when h >= a. Where, C0 is the nugget effect constant, C is the sill, and a is the range. Using optimization algorithms such as the least squares method to determine the parameters (C0, C, a) of the theoretical variogram model, so that the theoretical variogram curve best fits the experimental variogram points. 3. Anisotropy analysis: If the experimental variogram map shows obvious differences in different directions, it indicates that there is anisotropy in the spatial distribution of grades. It is necessary to determine the main range direction (the direction with the smallest variability) and the secondary range direction (the direction with the largest variability), and calculate the variogram parameters in different directions respectively. Repeat the above variogram analysis process for each key geological attribute of each semantic geological unit. Finally, obtain the variogram (VFMS), which contains the variogram models and their parameters of all semantic geological units and all key geological attributes. For each semantic geological unit, determine the key geological attributes (such as copper grade, ore body thickness) that need to be probabilistically evaluated. Then, using the borehole data in the integrated geological data, calculate the experimental variogram values of each key geological attribute, and select a suitable theoretical variogram model (such as the spherical model) for fitting to determine the parameters of the variogram model (nugget effect constant, sill, range) to obtain the variogram. Next, based on the variogram and the semantic geological unit, construct a Bayesian network structure, define the attribute nodes (key geological attributes), the reserve node, and the hidden nodes (such as blind faults), and determine the directed edges between the nodes. Determine the prior probability distribution of the nodes according to historical data and expert experience, and construct the conditional probability table of the attribute nodes using the variogram model and the Kriging interpolation method. Then, derive the conditional probability distribution of the reserve node through Monte Carlo simulation to obtain the unit probability data. Subsequently, if there is new exploration data, use the Gibbs sampling algorithm in the Markov chain Monte Carlo (MCMC) method for Bayesian inference to update the model parameters to obtain the updated probability data (if there is no new data, directly use the prior distribution). Finally, use the updated probability data for stochastic simulation to generate the probability distribution of the reserves of each semantic geological unit, and organize and normalize it to the unit reserve probability.

[0044] Step S3: Construct mining units for the semantic geological units to obtain mining units; construct an optimized mining objective function based on the mining units and the unit reserve probability to obtain a comprehensive objective function; construct a constraint function based on the mining units to obtain a constraint function; perform multi-objective optimization on the mining units according to the comprehensive objective function and the constraint function, and evaluate the mining plan to obtain the target mining plan;

[0045] In the embodiments of the present invention, according to the mining technical parameters (loader bucket capacity, blasting block size, minimum working space, etc.), the semantic geological unit is further divided into smaller mining units to obtain mining units. Then, a single-objective function for mining optimization (reserve maximization, grade balance, geological structure stability, mining efficiency) is defined and standardized. Next, the fuzzy comprehensive evaluation method is used to determine the weight of each single-objective function, and the standardized single-objective functions are weighted and summed to obtain a comprehensive objective function. At the same time, the mining constraint conditions of the mine (maximum slope angle, minimum mining width, maximum mining depth, safety distance, equipment capacity, etc.) are transformed into mathematical inequalities or equations, and the penalty function is constructed using the exterior point method to obtain the constraint function. Subsequently, taking the comprehensive objective function and the constraint function as inputs, the non-dominated sorting genetic algorithm II (NSGA-II) is used to perform multi-objective optimization on the mining units to obtain a set of Pareto-optimal mining sequence solution sets. Finally, the mining sequence solution set is transformed into a specific mining plan (including mining time step division, working face layout, equipment configuration, production plan, etc.), and these mining plans are evaluated (adding economic evaluation indicators such as net present value), the Pareto front is analyzed using tools such as parallel coordinate plots, and combined with the preferences of the manager, a target mining plan is selected.

[0046] Step S4: Simulate the mining process according to the target mining plan to obtain a mining performance data set; perform mining difference analysis on the mining performance data set to obtain a difference analysis report; correct the geological model according to the difference analysis report and the semantic geological unit to obtain a corrected geological model;

[0047] In the embodiments of the present invention, according to the target mining plan, semantic geological units, and mining units, a discrete event simulation method based on events is used to construct a mine mining simulation environment (including a three-dimensional geometric model, equipment model, event definition, and logical modeling). Then, the target mining plan is simulated for the mining process using this simulation environment, and various data generated during the simulation are recorded (such as the mining time, output, grade of each mining unit, the working time and utilization rate of equipment, etc.) to obtain the original simulation data record. Next, the original simulation data is processed and integrated (cleaned, transformed, and statistically analyzed) to generate a mining performance data set. Subsequently, the actual mining data of the mine is obtained, compared and analyzed with the mining performance data set to find the differences in aspects such as output, grade, and equipment utilization rate, and the reasons for the differences are analyzed to generate a difference analysis report. Then, according to the difference analysis report, the geological model parameters to be corrected and their correction directions are determined to obtain a model correction target set. Then, according to the model correction target, a suitable correction method (such as inverse distance weighted interpolation method, direct modification method, buffer method, etc.) is selected to adjust the model parameters of the semantic geological units to obtain an adjusted semantic unit set. Finally, model verification and evaluation (compared with the actual mining data) are performed based on the adjusted semantic unit set to generate a model verification report, and a final corrected geological model is formed according to the verification results.

[0048] Preferably, step S1 includes the following steps:

[0049] Step S11: Collect multi-source mine geological data; perform multi-source data preprocessing on the multi-source mine geological data to obtain fused geological data;

[0050] Step S12: Construct an initial three-dimensional geological model for the fused geological data and perform a preliminary grid division to obtain an initial geological grid model;

[0051] Step S13: Extract geological semantic features from the fused geological data according to the initial geological grid model to obtain a set of geological feature vectors;

[0052] Step S14: Perform geological semantic segmentation on the initial geological grid model according to the set of geological feature vectors to obtain an initial set of semantic units;

[0053] Step S15: Optimize the initial set of semantic units to obtain semantic geological units.

[0054] In the embodiments of the present invention, first, all available geological data within the mining area are comprehensively collected. These data include: borehole data, geophysical data, remote sensing data, existing geological maps, geological reports, exploration reports and other documentary materials in the mine. The collected multi-source heterogeneous geological data are preprocessed. Data cleaning: Check whether all data have missing values, outliers (such as grade values exceeding the reasonable range, coordinate errors, etc.) and duplicate values. Data format conversion: Uniformly convert data from different sources and different formats into a standard format. Coordinate system conversion and registration: Unify all data into the same coordinate system (such as the UTM projection coordinate system) and perform spatial registration. Data fusion (preliminary): For the geochemical analysis data (element grades) in the borehole data, use the Kriging interpolation method to generate a three-dimensional grade distribution map of the mining area based on the borehole sample point data. A fused geological data (FGD) containing all the preprocessed geological data is formed.

[0055] Using the information such as borehole data and geological maps in the fused geological data (FGD), an initial three-dimensional geological model of the ore deposit is constructed by using an implicit modeling method. 1. Stratum interface modeling: Based on the lithological stratification information in the borehole core logging data, use the radial basis function (RBF) interpolation algorithm to construct a three-dimensional surface of each stratum interface. The radial basis function interpolation takes the stratum interface points exposed by the boreholes as control points, and generates a continuous and smooth stratum interface surface by calculating the influence of each control point on any point in space. 2. Fault modeling: According to the fault information (fault location, attitude) exposed in the borehole data and the fault lines on the geological map, use the moving surface fitting algorithm to construct a three-dimensional surface of the fault plane. The moving surface fitting algorithm adjusts the surface shape to best fit the known fault data points. 3. Model combination: Perform Boolean operations (such as cutting, merging) on each stratum interface and fault plane to form a closed three-dimensional geological model, which represents the spatial distribution pattern of the ore body. Then, regular grid division is performed on the initial three-dimensional geological model. A three-dimensional block grid model is adopted to divide the space of the ore body and its surrounding rock into a series of cube units (or cuboid units) of the same size. The size of the grid unit is determined according to the minimum unit size of the mine exploitation and the calculation accuracy requirements. For example, it is set to 5m * 5m * 5m. Finally, an initial geological grid model (IGMM) is obtained.

[0056] Based on the initial geological grid model (IGMM) and the fused geological data (FGD), extract the geological semantic features of each grid cell and construct a geological feature vector set. 1. Geometric feature extraction: For each grid cell, calculate its center point coordinates (x, y, z), volume, shape index (such as aspect ratio, sphericity), and the distances to the nearest stratigraphic interface and fault plane. 2. Attribute feature extraction: Using the grade data in the fused geological data (FGD), adopt the inverse distance weighted interpolation method to calculate the average grade of each grid cell. For lithology, count the proportion of borehole samples of each lithology type within each grid cell, and take the lithology type with the largest proportion as the main lithology of the grid cell. Calculate the average density of the grid cell (estimated according to the lithology and its corresponding density value). 3. Topological feature extraction: Calculate the adjacency relationship between each grid cell and its adjacent grid cells (up, down, left, right, front, back), and construct an adjacency matrix. 4. Geological structure feature: For each grid cell, calculate its distance to the nearest fault and the influence degree of the fault on the grid cell (estimated according to the fault occurrence and rock mass strength). Using the geophysical data (taking electrical method data as an example) in the fused geological data, adopt a convolutional neural network (CNN), input the cross-sectional diagram of the electrical method data, and train a convolutional neural network model that can identify the electrical anomaly patterns related to mineralization. Use the trained CNN model to extract the deep geological features corresponding to each grid cell. Combine the geometric features, attribute features, topological features, and geological structure features of each grid cell into a feature vector. Finally, obtain a geological feature vector set (GFVS), which contains the geological semantic feature information of all grid cells.

[0057] Based on the geological feature vector set (GFVS) and the initial geological grid model (IGMM), use the K-means clustering algorithm to perform geological semantic segmentation on the grid cells. Data standardization: Standardize each feature in the geological feature vector set (GFVS) to eliminate the influence of dimension and numerical range differences. Clustering: Use the standardized feature vectors as the input of the K-means algorithm, and set the number of clusters K (determined according to the geological complexity of the ore deposit and the number of geological unit types, for example, K = 10). Boundary optimization: Use geological prior knowledge (such as stratigraphic contact relationship, fault cutting relationship) to optimize the clustering results. Isolated unit processing: For the isolated units (units without adjacent units of the same type) or too small units in the clustering results, merge them into the adjacent units with the most similar features.

[0058] Optimize the initial semantic unit set (ISUS) to generate semantic geological units (SGUS) that better conform to geological laws and mining practices. Unit merging: Merge units based on set rules. For example, if two adjacent semantic units have similar lithology, grade, and geological structure characteristics, and the shape of the merged unit is more regular, then these two units are merged into one unit. Unit splitting: Split semantic units with large internal property variations. For example, if the grade variation range within a semantic unit exceeds a preset threshold (e.g., the grade standard deviation is greater than 10%), then it is split into multiple sub-units with more uniform grades according to the grade distribution. Boundary adjustment: Adjust the boundaries of semantic units according to geological rules. For example, if the intersection line of the boundary of a semantic unit with a fault plane is not smooth, then the boundary is smoothed according to the strike and dip of the fault plane to make it more conform to the actual distribution pattern of the fault. Attribute review: For each optimized semantic unit, recalculate its comprehensive geological attributes (such as average grade, main lithology, density, etc.) to ensure the accuracy of attribute assignment. Manual inspection and adjustment: Manually inspect the optimized semantic unit set by geological engineers, and make manual adjustments to unreasonable unit divisions based on geological expertise and mining experience.

[0059] Step S21: Determine the key geological attributes of the semantic geological unit;

[0060] Step S22: Calculate the reserve probability of the semantic geological unit based on the Bayesian network structure according to the variogram to obtain unit probability data;

[0061] Step S23: Conduct Bayesian inference based on the unit probability data and update the parameters to obtain updated probability data;

[0062] Step S24: Generate the reserve probability distribution of the semantic geological unit based on the updated probability data to obtain the unit reserve distribution;

[0063] Step S25: Generate the unit reserve probability from the unit reserve distribution to obtain the unit reserve probability.

[0064] In the embodiments of the present invention, based on the Semantic Geological Unit (SGUS), for each semantic geological unit, key geological attributes that need to be probabilistically evaluated are determined. Based on the Variogram Function Model for Semantic Geological Units (VFMS) and the Semantic Geological Unit (SGUS), a Bayesian network structure is constructed for reserve probability calculation. 1. Node definition: For each semantic geological unit, an "attribute node" is created for each key geological attribute (such as copper grade, ore body thickness) to represent the probability distribution of the attribute. A "reserve node" is created to represent the reserve probability distribution of the unit. 2. Edge definition: A directed edge is added from each "attribute node" to the "reserve node" to represent the influence of the attribute on the reserve. According to geological knowledge, it is judged whether there is a dependency relationship between different attributes. If so, a directed edge is added between the corresponding "attribute nodes". For example, if high-grade ore bodies are usually accompanied by larger thicknesses, a directed edge is added between the "grade node" and the "thickness node". 3. Prior distribution determination: For each "attribute node", based on historical data (such as previous exploration data, data of similar ore deposits) and expert experience, its prior probability distribution is determined. For example, for copper grade, it can be assumed to follow a lognormal distribution, and the parameters (mean and variance) of the lognormal distribution are estimated according to historical data. For the "reserve node", its range can be initially estimated based on experience, and it is assumed to follow a uniform distribution. 4. Likelihood function construction: Using the variogram function model in the Variogram Function Model for Semantic Geological Units (VFMS), combined with the Kriging interpolation method, the conditional probability distribution of each "attribute node" is constructed. Specifically, for each semantic geological unit, using the Kriging interpolation method, based on the existing borehole data (from FGD) and the variogram function model, the attribute values and their estimated variances at unsampled locations within the unit are estimated. The mean estimated by Kriging is used as the expected value of the "attribute node", and the variance estimated by Kriging is used as the uncertainty measure of the "attribute node". According to the reserve calculation formula (such as the volume method: reserve = volume × density × grade), a functional relationship is established between the "reserve node" and other "attribute nodes", and combined with the conditional probability distribution of the "attribute node", the conditional probability distribution of the "reserve node" is derived. 5. Bayesian network construction: The nodes, edges, prior distributions, and likelihood functions are combined into a complete Bayesian network. For each semantic geological unit, the above Bayesian network construction process is repeated. Finally, unit probability data (UPMS) is obtained, which contains the Bayesian network models of all semantic geological units for reserve probability assessment.

[0065] Based on the unit probability data (UPMS) and new exploration data (if any), Bayesian inference is performed to update the model parameters. 1. Data preparation: If there is new exploration data (such as grades and thickness data of new boreholes), organize it into a format corresponding to the nodes of the Bayesian network. 2. Bayesian inference: Select the Gibbs sampling algorithm in the Markov chain Monte Carlo (MCMC) method for Bayesian inference. The Gibbs sampling algorithm approximates the posterior distribution of the entire network by iteratively sampling from the full conditional posterior distribution of each node. The specific steps are as follows: (1) Initialization: Set initial values for each node in the Bayesian network (which can be randomly sampled from the prior distribution). (2) Iterative sampling: For each node, fix the values of other nodes, and calculate the full conditional posterior distribution of this node according to its conditional probability distribution (determined by the Bayesian network structure and conditional probability table) and the observed data (if there is new data). Draw a new value from the full conditional posterior distribution and update the value of this node. Repeat this process until the preset number of iterations is reached. (3) Burn-in period: Discard the sampling results in the initial stage (burn-in period) because the sampling results in the initial stage may not have converged to the stationary distribution. (4) Sample collection: Collect samples from the sampling results after the burn-in period to form posterior samples. 3. Parameter update: Use the posterior samples to estimate the posterior probability distribution of each node. For the "reserve node", calculate statistics such as the expected value, variance, and confidence interval of the reserve according to its posterior samples. Replace the corresponding prior distribution in the unit probability data (UPMS) with the updated posterior distribution and update the model parameters.

[0066] If there is no new exploration data, skip the Bayesian inference step and directly use the prior distribution for subsequent reserve probability distribution generation. Finally, obtain the updated probability data (UPMS'), which contains the Bayesian network model updated (or not updated) by Bayesian inference.

[0067] Based on the updated probability data (UPMS') and semantic geological units (SGUS), generate the reserve probability distribution of each semantic geological unit.

[0068] Stochastic simulation: For each semantic geological unit, use its corresponding Bayesian network model (from UPMS') for stochastic simulation. Specifically, draw a set of random values from the posterior distribution (or prior distribution if Bayesian inference is not performed) of each node in the Bayesian network, and calculate the reserve value according to the reserve calculation formula (such as the volume method: reserve = volume × density × grade). Repeat this process multiple times (such as 10,000 times) to obtain a series of simulated values of the unit reserve.

[0069] Probability distribution estimation: Based on the reserve simulation values, draw a histogram or kernel density estimation curve of the reserves to obtain the probability distribution of the reserves. Calculate statistical quantities such as the expected value, variance, P10 (the value exceeded with a 10% probability), P50 (the median), and P90 (the value exceeded with a 90% probability) of the reserves to describe the uncertainty of the reserves.

[0070] Repeat the above random simulation and probability distribution estimation process for each semantic geological unit. Finally, obtain the Unit Reserve Distribution (URDS), which is a set containing the reserve probability distribution information (histogram, kernel density estimation curve, statistical quantities, etc.) of all semantic geological units.

[0071] Sort out and standardize the Unit Reserve Distribution (URDS) to generate the Unit Reserve Probability (URPS).

[0072] Data sorting: Sort out the data in the Unit Reserve Distribution (URDS) to ensure that the reserve probability distribution information of each semantic geological unit is stored in a unified format. For example, represent the reserve probability distribution of each unit as a table containing multiple quantiles (such as P10, P20,..., P90) and their corresponding reserve values, or store it as the parameters of a probability density function (PDF) if the probability distribution conforms to a certain known distribution, such as normal distribution, lognormal distribution, etc.

[0073] Data standardization: Perform standardization processing on the reserve data. For example, unify the reserve unit to tons and the grade unit to grams per ton or percentage.

[0074] Quality inspection: Check the integrity and consistency of the data to ensure that each semantic geological unit has corresponding reserve probability distribution information and that the data has no errors or omissions.

[0075] Set generation: Combine the reserve probability distribution information (after standardization) of all semantic geological units into a set to form the Unit Reserve Probability (URPS). This set stores the reserve uncertainty information of all units in a structured form and provides input for subsequent optimization of the mining sequence. URPS can be stored in the form of a database table, XML file, JSON file, etc.

[0076] Finally, obtain the Unit Reserve Probability (URPS), which stores the reserve probability distribution information of all semantic geological units in a standardized and structured form and provides key input for subsequent mining planning.

[0077] Preferably, step S22 includes the following steps:

[0078] Step S221: Define the attribute - reserve node of the semantic geological unit; Define the hidden node according to the fused geological data;

[0079] Step S222: Determine the Bayesian network structure for the attribute-reserve nodes and the hidden nodes to obtain the Bayesian network structure diagram;

[0080] Step S223: Obtain historical geological data; determine the prior probability distribution for the Bayesian network structure diagram based on the historical geological data to obtain the node prior probabilities;

[0081] Step S224: Determine the spatial correlation of geological attributes according to the variogram, and construct the likelihood function based on the Bayesian network structure diagram and the semantic geological unit to obtain the conditional probability table;

[0082] Step S225: Perform Bayesian inference on the Bayesian network structure diagram according to the node prior probabilities and the conditional probability table to obtain the node posterior probabilities;

[0083] Step S226: Extract the posterior probability distribution of the reserve nodes from the node posterior probabilities, and generate the unit probability model based on the Bayesian network structure diagram to obtain the unit probability data.

[0084] In the embodiment of the present invention, based on the semantic geological unit (SGUS), nodes of the Bayesian network are defined for each semantic geological unit. 1. Definition of attribute nodes: For each semantic geological unit, its key geological attributes (from step S21) are defined as "attribute nodes". For example, for a copper ore body unit, its copper grade, ore body thickness, density and other attributes are respectively defined as "copper grade node", "thickness node", "density node". 2. Definition of reserve nodes: For each semantic geological unit, a "reserve node" is created to represent the reserve of the unit. Combine all the "attribute nodes" and "reserve nodes" of all semantic geological units into a set, that is, the attribute-reserve nodes (ASNS). Then, according to the fused geological data (FGD) and geological prior knowledge, judge whether it is necessary to introduce hidden nodes. Hidden nodes represent geological factors that are not directly observed but affect the reserve. For example, if there is a buried fault in the mining area that is not fully exposed by drilling and this fault may affect the ore body grade and continuity, then a "buried fault node" can be created. The values of hidden nodes can be discrete (such as "exist", "not exist") or continuous (such as the distance from the fault to the ore body). Combine all the defined hidden nodes into a set, that is, the hidden nodes (HNS).

[0085] Based on the Attribute-Reserve Nodes (ASNS) and Hidden Nodes (HNS), determine the structure of the Bayesian network, i.e., the directed edges between nodes. 1. Edges between attribute nodes and reserve nodes: Add a directed edge from each "attribute node" to its corresponding "reserve node", indicating the direct influence of the attribute on the reserve. For example, add directed edges from the "copper grade node", "thickness node", and "density node" to the "reserve node" respectively. 2. Edges between attribute nodes: According to geological knowledge and data analysis, determine whether there is a dependency relationship between different attributes. If so, add a directed edge between the corresponding "attribute nodes". For example, if high-grade ore bodies are usually accompanied by larger thicknesses, add a directed edge between the "copper grade node" and the "thickness node" (the direction can be determined according to the causal relationship, such as from the "grade node" to the "thickness node"). 3. Edges between hidden nodes and other nodes: If there are hidden nodes, determine the relationship between the hidden nodes and other nodes according to geological knowledge. For example, if the presence or absence of the "buried fault node" affects the grade and continuity of the ore body, add directed edges from the "buried fault node" to the "copper grade node" and the "thickness node". The direction of the edge represents the causal relationship or conditional dependency relationship. Combine all the nodes and directed edges into a directed acyclic graph (DAG), which is the Bayesian network structure graph (BNSG). This graph represents the dependency relationship between variables and provides a basis for subsequent probability inference.

[0086] Collect historical geological data of the mining area and its surrounding areas, including previous exploration data, information on similar ore deposits, geological reports, scientific research literature, etc. These data are used to determine the prior probability distribution of the nodes in the Bayesian network. 1. Prior distribution of attribute nodes: For each "attribute node", determine the type and parameters of its prior probability distribution according to historical data and expert experience. For example, for the copper grade node, it can be assumed to follow a lognormal distribution, and the parameters (mean and variance) of the lognormal distribution can be estimated using historical data (such as grade data from previous drill holes). If the historical data is insufficient, data from similar ore deposits can be referred to or the opinions of geological experts can be consulted. For discrete attribute nodes (such as lithology), count the frequencies of different lithology types based on historical data as their prior probabilities. 2. Prior distribution of reserve nodes: For each "reserve node", preliminarily estimate its reserve range according to historical data and expert experience, and select an appropriate probability distribution to represent its uncertainty. For example, it can be assumed that the reserve follows a uniform distribution or a triangular distribution, and the parameters (lower limit, upper limit, most likely value) of the distribution can be determined according to experience. 3. Prior distribution of hidden nodes: For each "hidden node", determine its prior probability distribution according to geological knowledge and expert experience. For example, for the "buried fault node", its existence probability can be estimated based on the regional tectonic background and geophysical exploration data. Combine the prior probability distributions of all nodes into a set, which is the node prior probability (NPPS). This set provides initial information for Bayesian inference.

[0087] Based on the Variogram Function Model (VFMS), Bayesian Network Structure Graph (BNSG), and Semantic Geological Unit (SGUS), construct the likelihood function of the Bayesian network, that is, the conditional probability table. 1. Determination of spatial correlation of geological attributes: For each key geological attribute of each semantic geological unit, select the corresponding variogram function model from the Variogram Function Model (VFMS). This model describes the spatial variability and correlation of this attribute. 2. Kriging estimation: For each semantic geological unit, use the Kriging interpolation method to estimate the attribute values and their estimated variances at unsampled locations within the unit based on the existing borehole data (from FGD) and the variogram function model. The Kriging interpolation method takes into account spatial correlation and can provide an unbiased optimal estimate. 3. Construction of the conditional probability table for attribute nodes: For each "attribute node" in the Bayesian network, construct its conditional probability table (CPT) based on the results of Kriging estimation. If the node has no parent nodes, its CPT is its own probability distribution (which can be assumed to follow a normal distribution or other distribution according to the mean and variance of Kriging estimation). If the node has parent nodes, its CPT represents the probability distribution of the node given different values of the parent nodes. For example, if the "copper grade node" has a parent node "buried fault node", it is necessary to calculate the probability distribution of the "copper grade node" in the two cases of the existence and non-existence of the buried fault respectively. 4. Deduction of the conditional probability distribution of reserve nodes: For each "reserve node", deduce its conditional probability distribution according to the reserve calculation formula (such as the volume method) and the conditional probability tables of its parent nodes (attribute nodes). This can be achieved through Monte Carlo simulation: randomly draw a set of values from the conditional probability distributions of each parent node, calculate the reserve value according to the reserve calculation formula, repeat multiple times to obtain a series of simulated values of the reserve, so as to estimate the conditional probability distribution of the reserve. Combine the conditional probability tables of all nodes into a set, that is, the conditional probability table set (CPTS). This set represents the probability of the child node taking values given the values of the parent nodes and is the core component of the Bayesian network.

[0088] Based on the node prior probability (NPPS), conditional probability tables (CPTs), and Bayesian network structure graph (BNSG), Bayesian inference is carried out to calculate the posterior probability distribution of each node. Select a suitable Bayesian inference algorithm. Since this is a static Bayesian network and the number of nodes and states is limited, an exact inference algorithm such as variable elimination can be used. The basic idea of variable elimination is: by sequentially eliminating the non-evidence nodes (non-observed nodes) in the network, the joint probability distribution is decomposed into a product of a series of conditional probabilities, thereby calculating the posterior probability of the target node (the node for which the posterior probability needs to be calculated). The specific steps include: 1. Determine the order of variable elimination. A good elimination order can reduce the computational complexity. 2. For each variable to be eliminated, multiply all the factors (conditional probability tables) where it is located to obtain a new factor. 3. Sum (or integrate) over this variable in the new factor to eliminate it. 4. Repeat steps 2 and 3 until all non-evidence nodes are eliminated. 5. Multiply the remaining factors and normalize them to obtain the posterior probability distribution of the target node. If the network structure is complex and the number of nodes and states is large, approximate inference algorithms such as variational inference or Markov chain Monte Carlo (MCMC) methods can be used. Calculate the posterior probability distribution of each node (including attribute nodes, reserve nodes, and hidden nodes). Combine the posterior probability distributions of all nodes into a set, namely the node posterior probability (NPPS'').

[0089] Based on the node posterior probability (NPPS'') and the Bayesian network structure graph (BNSG), extract the reserve probability model of each semantic geological unit. 1. Extraction of reserve node posterior probability distribution: From the node posterior probability (NPPS''), extract the posterior probability distribution of the "reserve node" corresponding to each semantic geological unit. This distribution reflects the updated understanding of the reserve of this unit after considering geological prior knowledge, attribute data, and the Bayesian network structure. 2. Generation of unit probability model: Process the posterior probability distribution of the "reserve node" of each semantic geological unit to generate the probability model of this unit. Specifically, it includes: calculating statistical quantities such as the expected value, variance, and confidence interval of the reserve to quantitatively describe the uncertainty of the reserve. Plot the probability density function (PDF) or cumulative distribution function (CDF) curve of the reserve for visualizing the probability distribution of the reserve. If the posterior probability distribution conforms to a certain known distribution (such as normal distribution, lognormal distribution, etc.), its distribution parameters can be estimated. 3. Generation of model set: Combine the probability models (including statistical quantities, probability distribution curves, or distribution parameters) of all semantic geological units into a set, namely the unit probability data (UPMS). This set stores the reserve probability models of all units in a structured form, providing key inputs for subsequent mining planning. Finally, obtain the unit probability data (UPMS), which contains the reserve probability models of all semantic geological units and provides a basis for subsequent optimization of the mining sequence.

[0090] Preferably, step S224 is specifically as follows:

[0091] Extract the geological attributes of the semantic geological units, calculate the correlation coefficients of the attributes between different spatial positions based on the variogram function, and obtain the attribute spatial correlation;

[0092] Perform Kriging estimation based on the attribute spatial correlation and the semantic geological units to obtain the estimated values of the attribute nodes;

[0093] Construct the conditional probability table of the attribute nodes for the Bayesian network structure diagram based on the estimated values of the attribute nodes to obtain the conditional probability table of the attribute nodes;

[0094] Derive the conditional probability distribution of the reserve nodes based on the conditional probability table of the attribute nodes to obtain the conditional probability table.

[0095] In the embodiment of the present invention, based on the semantic geological unit (SGUS), the key geological attribute data of each semantic geological unit is extracted. For example, for a copper ore body unit, the copper grade values of all drill hole data inside it are extracted. Then, according to the variogram function (VFMS), the variogram function model corresponding to the key geological attribute of each semantic geological unit is selected. For example, if the copper grade variogram function model of this copper ore body unit recorded in VFMS is the spherical model, and the parameters are: nugget effect C0 = 0.01, sill C = 0.05, range a = 100 meters, then this model is used to calculate the correlation coefficients of the copper grade between different spatial positions. The specific calculation process is as follows: For any two position points xi and xj inside this copper ore body unit, calculate the distance h = ||xi - xj|| between them. According to the spherical model formula, calculate the variogram value γ(h) = 0.01 + 0.05×[1.5(h / 100) - 0.5(h / 100)^3] (when h < 100 meters) or γ(h) = 0.01 + 0.05 = 0.06 (when h >= 100 meters). Then, calculate the correlation coefficient ρ(h) = (C0 + C - γ(h)) / (C0 + C) = 1 - γ(h) / 0.06 between these two positions. Repeat the above calculation process for each key geological attribute of each semantic geological unit. Finally, obtain the attribute spatial correlation (ASCM), and this set contains the correlation coefficient calculation models (i.e., the variogram function model and its parameters) of each semantic geological unit and each key geological attribute between different spatial positions.

[0096] Based on the Attribute Space Correlation (ASCM) and Semantic Geological Unit (SGUS), Kriging estimation is carried out for each key geological attribute of each semantic geological unit. Taking ordinary Kriging as an example: for a copper ore body unit, estimate the copper grade value at an unsampled location x0 inside it. 1. Determine the neighborhood: According to the range of the variogram model of the copper grade in this unit in ASCM (for example, 100 meters), determine the neighborhood range of x0, and select the borehole data points with known copper grade values within this range (assuming there are n borehole data points, with their positions being x1, x2,..., xn, and the corresponding copper grade values being Z(x1), Z(x2),..., Z(xn)). 2. Construct the Kriging equations: According to the principle of ordinary Kriging, construct the Kriging equations:

[0097] Σλi×C(xi, xj)+μ = C(x0, xj) (j = 1, 2,..., n); Σλi = 1;

[0098] where λi is the weight coefficient, C(xi, xj) is the covariance between positions xi and xj (calculated according to the variogram model in ASCM, C(h)=C0 + C-γ(h)), and μ is the Lagrange multiplier. 3. Solve the Kriging equations: Use matrix operations to solve the above equations to obtain the weight coefficient λi and the Lagrange multiplier μ. 4. Calculate the estimated value and variance: Calculate the estimated copper grade value at x0, Z×(x0)=Σλi×Z(xi). Calculate the estimated variance σ^2(x0)=Σλi×C(x0, xi)+μ. For each key geological attribute of each semantic geological unit, repeat the above Kriging estimation process at all unsampled positions inside it (the grid can be divided at a certain interval). Finally, obtain the Attribute Node Estimated Values (AEVS), which is a set containing the Kriging estimated values and estimated variances of each semantic geological unit, each key geological attribute at each unsampled position.

[0099] Based on the Attribute Node Estimation Value (AEVS) and the Bayesian Network Structure Graph (BNSG), construct the Conditional Probability Table (CPT) of the "attribute nodes" in the Bayesian network. For the "copper grade node" of a copper ore body unit, if the node has no parent node, its CPT is its own probability distribution. According to the Kriging estimation value and the estimation variance of the copper grade of this unit in the AEVS, assume that the copper grade follows a normal distribution, and use the Kriging estimation value as the mean of the normal distribution and the Kriging estimation variance as the variance of the normal distribution, so as to determine the CPT of this node. If the "copper grade node" has a parent node (such as the "buried fault node"), its CPT represents the probability distribution of the "copper grade node" given different values of the parent node. For example, if the "buried fault node" has two values: "exists" and "does not exist", it is necessary to calculate the probability distribution of the "copper grade node" in the two cases of the existence and non-existence of the buried fault respectively (also assume a normal distribution and determine the parameters of the normal distribution according to the Kriging estimation results in different cases). For other "attribute nodes" (such as the thickness node, density node), repeat the above CPT construction process. Finally, obtain the Attribute Node Conditional Probability Tables (ACPTS), which is a set containing the conditional probability tables of all "attribute nodes" in the Bayesian network.

[0100] Based on the Attribute Node Conditional Probability Tables (ACPTS), deduce the conditional probability distribution of the "reserves node" in the Bayesian network. For the "reserves node" of a copper ore body unit, its reserves are jointly determined by the values of its parent nodes (attribute nodes, such as the "copper grade node", "thickness node", "density node") and the reserves calculation formula (such as the volume method: reserves = volume × density × grade). Use the Monte Carlo simulation method to deduce the conditional probability distribution of the "reserves node": 1. From the conditional probability tables of each "attribute node" in the ACPTS, randomly draw a set of attribute values (such as a set of copper grade values, thickness values, and density values) according to the values of its parent nodes (if there are parent nodes). 2. According to the reserves calculation formula, calculate the reserves value using the drawn attribute values. 3. Repeat steps 1 and 2 multiple times (such as 10,000 times) to obtain a series of simulation values of the reserves. 4. According to the reserves simulation values, draw a histogram or kernel density estimation curve of the reserves to obtain the conditional probability distribution of the "reserves node". It is also possible to divide the reserves values into several intervals and count the frequency of each interval to obtain a discrete form of the conditional probability table.

[0101] For the "reserves node" of other semantic geological units, repeat the above process of deducing the conditional probability distribution. Finally, obtain the Conditional Probability Tables (CPTS), which is a set containing the conditional probability tables of all nodes (including attribute nodes and reserves nodes) in the Bayesian network.

[0102] Preferably, step S23 includes the following steps:

[0103] Step S231: Obtain newly added exploration data; perform integration of observation data on the newly added exploration data to obtain integrated observation data;

[0104] Step S232: Perform Gibbs sampling on the Bayesian network structure diagram according to the conditional probability table and the integrated observation data to obtain posterior samples;

[0105] Step S233: Calculate the posterior distribution based on the posterior samples to obtain node posterior probabilities;

[0106] Step S234: Update the graph parameters of the Bayesian network structure diagram using the node posterior probabilities to obtain updated probability data.

[0107] In the embodiment of the present invention, the newly generated exploration data after the reserve probability assessment of the mine is obtained, including newly drilled borehole data (lithology, grade, thickness, etc.), actual geological data (lithology, grade, structure, etc.) exposed by the newly mined mining face, new geophysical exploration data, etc. The newly added exploration data is preprocessed and integrated to correspond to the nodes in the Bayesian network. The integrated observation data (IOD) is obtained.

[0108] Based on the conditional probability tables (CPTs), integrated observed data (IOD), and Bayesian network structure graph (BNSG), the Gibbs Sampling algorithm is used for Bayesian inference to generate posterior samples. Gibbs Sampling is a Markov Chain Monte Carlo (MCMC) method that approximates the joint posterior distribution by iteratively sampling from the full conditional posterior distribution of each variable. The specific steps are as follows: 1. Initialization: Set an initial value for each node in the Bayesian network. It can be randomly drawn from the prior distribution of the node or set to any reasonable value. 2. Iterative sampling: Sample each node in turn according to a predetermined order (or a random order). For the current node to be sampled (assumed to be node Xi): Fix the values of all other nodes (denoted as X-i). According to the Bayesian network structure and conditional probability tables (CPTs), determine the full conditional posterior distribution of node Xi, that is, P(Xi|X-i, IOD). This distribution depends only on the nodes in the Markov Blanket of Xi and the observed data IOD. Draw a new value from P(Xi|X-i, IOD) as the updated value of node Xi. 3. Repeat step 2 for all nodes for one round of sampling, which is called one iteration. 4. Repeat the iteration multiple times (e.g., 10,000 times). 5. Burn-in period: Discard the sampling results in the initial stage (e.g., the first 1,000 iterations) because these samples may not have converged to the stationary distribution. 6. Sample collection: After the burn-in period, collect the values of all nodes at regular intervals (e.g., every 10th iteration) to form a sample. Combine all the collected samples into a set, which is the posterior sample set (PSS).

[0109] Based on the posterior samples (PSS), calculate the posterior probability distribution of each node in the Bayesian network. For each node: Continuous nodes (such as grade, thickness, reserves): Use the sample values of this node in the posterior samples (PSS) and adopt the Kernel Density Estimation (KDE) method to estimate its probability density function (PDF). KDE is a non-parametric method. By placing a kernel function (such as a Gaussian kernel function) at each sample point and then superimposing all the kernel functions, an estimate of the probability density function is obtained. Statistical quantities such as the sample mean, variance, and quantiles can also be calculated as the characteristics of the posterior distribution. Discrete nodes (such as lithology, presence or absence of blind faults): Count the frequencies of each value of this node in the posterior samples (PSS) as the estimate of its posterior probability. For example, if the "blind fault node" has two values, "present" and "absent", and in 1000 posterior samples, "present" appears 600 times and "absent" appears 400 times, then the posterior probability estimate of "present" is 0.6, and the posterior probability estimate of "absent" is 0.4. Combine the posterior probability distributions (PDF or discrete probabilities) of all nodes into a set, namely the node posterior probability (NPPS'').

[0110] Based on the node posterior probability (NPPS''), update the parameters of the Bayesian network model to obtain updated probability data. The specific operations include: 1. Update the conditional probability table: Replace the corresponding conditional probability distribution in the conditional probability table (CPTS) with the posterior probability distribution in the node posterior probability (NPPS''). For example, if the posterior distribution of the "copper grade node" is a normal distribution N(μ', σ'^2), then update the conditional probability table of this node in the CPTS to this normal distribution. If the posterior probability of the "blind fault node" is P(present)=0.6 and P(absent)=0.4, then update the conditional probability table of this node in the CPTS to these two probability values. 2. Update the reserve probability distribution: For the "reserve node" of each semantic geological unit, its posterior probability distribution is already included in NPPS''. According to the new posterior distribution, recalculate statistical quantities such as the expected value, variance, and confidence interval of the reserves, and update the probability density function (PDF) or cumulative distribution function (CDF) of the reserves. 3. Update the model set: Integrate the updated conditional probability table and reserve probability distribution information into the unit probability data (UPMS) to form the updated probability data (UPMS'). Finally, obtain the updated probability data (UPMS').

[0111] Preferably, step S3 includes the following steps:

[0112] Step S31: Obtain the mine mining technical parameters and mining constraint conditions; construct mining units for the semantic geological units according to the mine mining technical parameters to obtain mining units;

[0113] Step S32: Construct an exploitation optimization objective function based on the exploitation unit and the unit reserve probability to obtain a comprehensive objective function;

[0114] Step S33: Convert the exploitation constraint conditions into mathematical expressions, and construct a constraint function based on the exploitation unit and the mine exploitation technical parameters to obtain a constraint function;

[0115] Step S34: Use the comprehensive objective function and the constraint function as the input of the optimization algorithm, and perform multi-objective optimization solution for the exploitation unit based on the multi-objective optimization algorithm to obtain an exploitation sequence solution set;

[0116] Step S35: Generate an exploitation plan from the exploitation sequence solution set to obtain an exploitation plan;

[0117] Step S36: Evaluate the exploitation plan and perform Pareto optimal solution analysis to obtain the target exploitation plan.

[0118] In the embodiment of the present invention, the mine exploitation technical parameters and exploitation constraint conditions actually adopted are obtained. The exploitation technical parameters include: the model, rated bucket capacity, loading capacity, and average loading time of the loader; the model, load capacity, average driving speed, and unloading time of the transport truck; the blasting parameters, such as the type of explosive, single-hole charge amount, blasting network design, and maximum charge amount of the largest section initiation; the advancing speed of the mining working face, etc. The exploitation constraint conditions include: the maximum slope angle (such as 45 degrees), the minimum exploitation width (such as 20 meters), the maximum exploitation depth (such as altitude -500 meters), the safety distance (such as more than 50 meters away from the fault), etc. According to the mine exploitation technical parameters, the semantic geological unit (SGUS) is further divided into smaller exploitation units suitable for actual exploitation. The specific operations include: 1. Unit refinement: According to the bucket capacity of the loader and the size of the blasted block, a larger semantic geological unit (such as the entire ore body) is further divided into smaller blocks. For example, if the bucket capacity of the loader is 10 cubic meters and the size of the blasted block designed is 5 m × 5 m × 5 m, the semantic geological unit can be refined according to the size of 5 m × 5 m × 5 m. 2. Unit merging: For too small semantic geological units (such as isolated and very thin intercalated rock layers), according to the minimum operation space requirements of the exploitation equipment, they are merged with adjacent units. For example, if the minimum operation width of the loader is 3 meters, the thin intercalated rock layer with a width less than 3 meters is merged with the ore body unit of its hanging wall or footwall. 3. Unit adjustment: According to the advancing direction and shape of the mining working face, the boundary of the exploitation unit is adjusted to make it more conform to the geometric shape of actual exploitation. For example, the boundary of the exploitation unit is adjusted to an inclined plane parallel to the mining working face instead of a vertical plane. 4. Unit numbering: All the divided exploitation units are numbered uniquely. Finally, the exploitation unit (MUS) is obtained.

[0119] Based on the Mining Unit (MUS) and Unit Reserve Probability (URPS), an objective function for optimizing the mining sequence is constructed. The objective function needs to comprehensively consider multiple objectives of mine mining, such as reserve maximization, grade balance, geological structure stability, mining efficiency, etc. The specific operations include:

[0120] 1. Define single-objective functions:

[0121] Reserve maximization (f1): For a given mining sequence, calculate the sum of the expected reserves of all mining units. The expected reserve can be calculated according to the reserve probability distribution of each unit in the Unit Reserve Probability (URPS). The objective function f1 is the opposite of the total reserve (because usually the optimization algorithm minimizes the objective function). Grade balance (f2): For a given mining sequence, calculate the variance of the expected grades (according to URPS) of all mining units. The objective function f2 is this variance. Geological structure stability (f3): According to the geological structure information (such as faults, fractures, etc.) in the Semantic Geological Unit (SGUS), evaluate the geological structure stability for each mining unit. A stability scoring function can be defined. For example, the farther away from the fault and the better the rock mass integrity, the higher the score. The objective function f3 is the opposite of the average value of the stability scores of all units in the mining sequence. Mining efficiency (f4): According to the geometric characteristics of the mining units (such as location, volume) and the mine mining technical parameters (such as the efficiency of loaders and haul trucks), estimate the mining time and transportation time of each mining unit. The objective function f4 can be the total mining time, total transportation time of the mining sequence, or their weighted sum.

[0122] 2. Standardize the objective functions: Since the dimensions and value ranges of different objective functions are different, they need to be standardized to make them comparable. The range normalization method can be used to map the values of each objective function to the interval [0, 1].

[0123] 3. Determine weights: Determine the weights of each objective function to reflect its importance in the comprehensive objective function. The Analytic Hierarchy Process (AHP) or the expert scoring method can be used to determine the weights.

[0124] 4. Construct the comprehensive objective function: Multiply the standardized single-objective functions by their corresponding weights and then sum them up to obtain the comprehensive objective function (COF). For example: COF = w1×f1' + w2×f2' + w3×f3' + w4×f4', where w1, w2, w3, w4 are the weights of the four objective functions respectively, and f1', f2', f3', f4' are the standardized objective function values respectively.

[0125] Convert the mining constraint conditions of the mine into mathematical expressions and construct constraint functions. The specific operations include:

[0126] 1. Classification of Constraints: The mining constraints are classified into geometric constraints, production constraints, and geological constraints.

[0127] Geometric constraints: such as the maximum slope angle, the minimum mining width, and the maximum mining depth. Production constraints: such as the loading capacity of the loader and the transportation capacity of the haul truck. Geological constraints: such as the safety distance from the fault.

[0128] 2. Mathematical Expression of Constraints: Each constraint is transformed into a mathematical inequality or equation.

[0129] Maximum slope angle constraint: Assume that mining units i and j are adjacent vertically, with i above and j below. Let the height difference between them be Δh, the horizontal distance be Δd, and the maximum slope angle be α. Then the constraint can be expressed as: arctan(Δh / Δd) <= α. Minimum mining width constraint: Let the width of mining unit i be wi and the minimum mining width be Wmin. Then the constraint can be expressed as: wi >= Wmin. Maximum mining depth constraint: Let the bottom elevation of mining unit i be zi and the maximum mining depth be Zmax. Then the constraint can be expressed as: zi >= Zmax. Loader loading capacity constraint: Let the combined mining units mined in each time period be S, and the maximum loading capacity of the loader be Qmax. Then the constraint can be expressed as: Σqi <= Qmax (i ∈ S), where qi is the ore quantity of mining unit i. Haul truck transportation capacity constraint: Let the combined mining units transported in each time period be S, and the maximum transportation capacity of the haul truck be Tmax. Then the constraint can be expressed as: Σqi <= Tmax (i ∈ S). Safety distance constraint from the fault: Let the closest distance from mining unit i to the fault be di and the safety distance be Dsafe. Then the constraint can be expressed as: di >= Dsafe.

[0130] 3. Construction of Constraint Function: Using the exterior penalty function method, the constraints are transformed into penalty terms of the objective function. For the inequality constraint g(x) <= 0, the penalty term is P(x) = α × max(0, g(x))^2, where α is the penalty factor. For the equality constraint h(x) = 0, the penalty term is P(x) = β × h(x)^2, where β is the penalty factor. Add all the penalty terms to obtain a total penalty function as the constraint function.

[0131] Taking the comprehensive objective function constructed in step S32 and the constraint function constructed in step S33 as inputs, the NSGA-II algorithm is used for solving. Set the population size to 100, the number of iterations to 500, the crossover probability to 0.9, and the mutation probability to 0.1. The NSGA-II algorithm first initializes a population containing 100 random mining sequences. Each mining sequence represents the mining order of mining units. The algorithm ranks the individuals in the population through non-dominated sorting and crowding distance calculation, selects excellent individuals for crossover and mutation operations, and generates new mining sequences. The crossover operation exchanges some segments of two parent mining sequences to generate two new offspring sequences. The mutation operation randomly changes the mining order of some units in the mining sequence. Iterate 500 times in this way, and finally output a set containing 100 Pareto optimal solutions, that is, the solution set of mining sequences. Each Pareto optimal solution represents a mining sequence that cannot be dominated by other solutions in terms of the comprehensive objective function. For example, one solution may perform best in terms of maximizing reserves, while another solution may perform best in terms of geological structure stability.

[0132] Traverse each mining sequence in the solution set of mining sequences obtained in step S34. Taking one of the mining sequences as an example, this sequence defines the mining order of 1000 mining units. According to the volume of each mining unit and the preset blasting parameters (for example, the charge per hole is 2 kg / m³ and the explosive type is emulsion explosive), calculate the amount of explosive and blasting time required for each mining unit. Assume that the average blasting time for each mining unit is 2 hours. Combining the parameters of the existing loaders in the mine (for example, 3 loaders with a bucket capacity of 10 m³ and an average loading time of 5 minutes per bucket) and transport trucks (for example, 5 trucks with a load capacity of 50 tons and an average transport time of 30 minutes per trip), arrange the operation plans for the loaders and transport trucks. For example, divide the 1000 mining units into 100 batches, with each batch containing 10 mining units. The mining process for each batch is as follows: First, perform the blasting operation, which takes 2 hours. Then, 3 loaders load the ore simultaneously. Each loader needs to load approximately 34 buckets of ore (the total volume of 10 units / bucket capacity), which takes approximately 170 minutes. 5 transport trucks transport the ore in a cycle. Each truck transports 50 tons of ore each time, and calculate the number of transports according to the ore density and volume of each mining unit. Assume that each truck needs to transport 5 times on average, then the total transport time is approximately 150 minutes. Therefore, the mining time for each batch is approximately 2 + 170 / 60 + 150 / 60 = 7.33 hours. Arrange the mining plans for the 100 batches in order to form a complete mining plan, including information such as the specific mining time of each mining unit, the required equipment and personnel arrangements, blasting parameters, and transport routes.

[0133] Evaluate each mining plan generated in step S35. Taking the net present value (NPV) and the geological structure stability score as examples for illustration. Calculate the net present value of each mining plan according to the ore volume, grade, mining cost (such as explosive cost, equipment depreciation, labor cost, etc.) and metal price of each mining unit in each mining plan. Calculate the geological structure stability score of each mining unit according to factors such as the distance of each mining unit from the fault and the rock mass quality, and take the average value of the scores of all mining units as the geological structure stability score of this mining plan. Plot the net present value and the geological structure stability score of all mining plans in a two-dimensional coordinate system to form a Pareto frontier graph. The points on the Pareto frontier graph represent the mining plans corresponding to the Pareto optimal solutions. By analyzing the Pareto frontier graph, a mining plan that balances different objectives can be found. For example, a certain plan has the highest net present value but a lower geological structure stability score; while another plan has a slightly lower net present value but a higher geological structure stability score. Select the final target mining plan according to the actual situation and risk preference of the mine. For example, if the mine focuses on economic benefits, the plan with the highest net present value can be selected; if the mine focuses on safety in production, the plan with the highest geological structure stability score can be selected.

[0134] Preferably, step S32 includes the following steps:

[0135] Step S321: Define a single-objective function according to the mining unit and the unit reserve probability to obtain a single-objective function, where the single-objective function includes the objectives of maximizing reserves, grade balance, geological structure stability, and mining efficiency;

[0136] Step S322: Standardize the single-objective function to obtain a standardized single-objective function;

[0137] Step S323: Determine the objective weight based on fuzzy comprehensive evaluation according to the standardized single-objective function to obtain an objective weight vector;

[0138] Step S324: Construct a comprehensive objective function according to the standardized single-objective function and the objective weight vector to obtain a comprehensive objective function.

[0139] In the embodiments of the present invention, four single-objective functions are defined based on the mining unit (MUS) and the unit reserve probability (URPS). The total number of mining units is 50, numbered from 1 to 50. The reserve probability distribution of each mining unit is known and represented in a discrete form, including three quantiles P10, P50, P90 and their corresponding reserve values (unit: ton). The copper grade probability distribution of each mining unit is known and represented in a discrete form, including three quantiles P10, P50, P90 and their corresponding copper grade values (unit: %). The distance from each unit to the fault is known (unit: meter). 1. Reserve maximization objective (f1): For a given mining sequence (a permutation of 50 mining units), calculate its total expected reserve. The expected reserve of each mining unit is the reserve value corresponding to its P50 quantile. The total expected reserve of the mining sequence is the sum of the expected reserves of all mining units. The objective function f1 is the opposite of the total expected reserve (because optimization algorithms usually minimize the objective function). 2. Grade balance objective (f2): For a given mining sequence, calculate the variance of the mined copper grade. First, calculate the expected copper grade of each mining unit, which is the copper grade value corresponding to its P50 quantile. Then, calculate the variance of the expected copper grades of all mining units in the mining sequence. The objective function f2 is this variance. 3. Geological structure stability objective (f3): For a given mining sequence, calculate its average distance to the fault. The distance from each mining unit to the fault is known. The average distance to the fault of the mining sequence is the average of the distances from all mining units to the fault. The objective function f3 is the opposite of the average distance to the fault. 4. Mining efficiency objective (f4): For a given mining sequence, calculate its total mining time. Assume that the mucking loader can handle 100 cubic meters of ore per hour and the transport truck can transport 50 cubic meters of ore per hour. The volume of each mining unit is known (calculated based on its geometry and dimensions). The mining time of a mining unit is its volume divided by the handling speed of the mucking loader. The transport time of a mining unit is its volume divided by the transport speed of the transport truck. The total mining time of the mining sequence is the sum of the mining times and transport times of all mining units. The objective function f4 is the total mining time. Finally, the single-objective functions are obtained, including f1, f2, f3 and f4.

[0140] Standardize the single-objective function obtained in step S321. The range standardization method is adopted. Assume that through the calculation of a large number of randomly generated mining sequences, the following data are obtained: The maximum value (-f1_min, that is, the minimum total expected reserve) of the reserve maximization objective (f1) is -10 million tons, and the minimum value (-f1_max, that is, the maximum total expected reserve) is -6 million tons. The maximum value (f2_max) of the grade balance objective (f2) is 0.1, and the minimum value (f2_min) is 0.01. The maximum value (-f3_min) of the geological structure stability objective (f3) is -50 meters, and the minimum value (-f3_max) is -200 meters. The maximum value (f4_max) of the mining efficiency objective (f4) is 50,000 hours, and the minimum value (f4_min) is 30,000 hours.

[0141] For a given mining sequence, the values of its four single-objective functions are: f1 = -8 million tons, f2 = 0.05, f3 = -100 meters, f4 = 40,000 hours. Then the standardized single-objective function values are:

[0142] f1'=(f1 - f1_min) / (f1_max - f1_min)=(-800 - (-1000)) / (-600 - (-1000)) = 200 / 400 = 0.5;

[0143] f2'=(f2 - f2_min) / (f2_max - f2_min)=(0.05 - 0.01) / (0.1 - 0.01)=0.04 / 0.09 = 0.444;

[0144] f3'=(f3 - f3_min) / (f3_max - f3_min)=(-100 - (-50)) / (-200 - (-50))=-50 / -150 = 0.333;

[0145] f4'=(f4 - f4_min) / (f4_max - f4_min)=(40000 - 30000) / (50000 - 30000)=10000 / 20000 = 0.5;

[0146] Finally, the standardized single-objective function is obtained, including f1', f2', f3', and f4'.

[0147] According to the standardized single-objective function, the fuzzy comprehensive evaluation method is used to determine the objective weight vector.

[0148] 1. Establish a factor set: The factor set U = {u1, u2, u3, u4}, corresponding to the four objectives of reserve maximization, grade balance, geological structure stability, and mining efficiency respectively.

[0149] 2. Establish the evaluation set: The evaluation set V = {v1, v2, v3, v4, v5} = {very important, important, average, less important, very unimportant}.

[0150] 3. Establish the fuzzy relation matrix: Invite 5 experts to evaluate the importance of four objectives. Each expert gives the membership degree of each objective relative to each evaluation in the evaluation set. For example, for objective u1 (maximizing reserves), an expert gives the following membership degrees: (0.6, 0.3, 0.1, 0, 0), indicating that the membership degree of the objective of maximizing reserves being "very important" is 0.6, "important" is 0.3, "average" is 0.1, and the membership degrees of "less important" and "very unimportant" are 0. Combine the evaluation results of the 5 experts to obtain the fuzzy relation matrix R:

[0151] R = | 0.6 0.3 0.1 0.0 0.0 | (u1)

[0152] | 0.2 0.4 0.3 0.1 0.0 | (u2)

[0153] | 0.3 0.4 0.2 0.1 0.0 | (u3)

[0154] | 0.1 0.3 0.4 0.2 0.0 | (u4)

[0155] Each row represents an objective, and each column represents an evaluation level.

[0156] Determine the weight vector: Normalize the fuzzy relation matrix R so that the sum of the elements in each row is 1. Then, calculate the weight of each objective. The weight calculation formula is: wi = (Σrij) / (ΣΣrij), where rij is the element in the i-th row and j-th column of the fuzzy relation matrix R.

[0157] R_normalized = | 0.6 0.3 0.1 0.0 0.0 |

[0158] | 0.2 0.4 0.3 0.1 0.0 |

[0159] | 0.3 0.4 0.2 0.1 0.0 |

[0160] | 0.1 0.3 0.4 0.2 0.0 |

[0161] The calculated weight vector W = (0.4, 0.2, 0.2, 0.2).

[0162] Construct a comprehensive objective function based on the standardized single-objective function and the objective weight vector. Assume that the standardized single-objective function values of a mining sequence are: f1' = 0.5, f2' = 0.444, f3' = 0.333, f4' = 0.5. The objective weight vector is W = (0.4, 0.2, 0.2, 0.2).

[0163] Then the comprehensive objective function value is:

[0164] COF = w1×f1' + w2×f2' + w3×f3' + w4×f4' = 0.4×0.5 + 0.2×0.444 + 0.2×0.333 + 0.2×0.5 = 0.2 + 0.0888 + 0.0666 + 0.1 = 0.4554;

[0165] The comprehensive objective function value (COF) of this mining sequence is 0.4554.

[0166] The comprehensive objective function (COF) performs a weighted sum of the four standardized single-objective function values. The weights reflect the relative importance of each single-objective function in the comprehensive evaluation. By adjusting the weights, the trade-off relationship between different objectives can be changed, thus affecting the final optimization result. In practical applications, the weights need to be determined according to the preferences and decision-making objectives of the mine managers. The finally constructed comprehensive objective function (COF) is a scalar function, which transforms the multi-objective optimization problem into a single-objective optimization problem, providing an objective function for subsequent mining sequence optimization algorithms (such as NSGA-II).

[0167] Preferably, step S33 includes the following steps:

[0168] Step S331: Classify the mining constraint conditions to obtain a classified constraint set, where the classified constraint set includes a geometric constraint set, a production constraint set, and a geological constraint set;

[0169] Step S332: Express the classified constraint set mathematically to obtain a constraint expression;

[0170] Step S333: Select an exterior point method penalty function for the constraint expression and determine the penalty factor to obtain a penalty function;

[0171] Step S334: Generate a constraint function from the penalty function to obtain a constraint function.

[0172] In the embodiments of the present invention, the constraint conditions for mine exploitation are classified. The exploitation units are 50 in number, numbered from 1 to 50. 1. Geometric constraints: Maximum slope angle constraint: The final slope angle formed by exploitation shall not exceed 45 degrees. Minimum exploitation width constraint: The exploitation width of each exploitation unit shall not be less than 20 meters. Maximum exploitation depth constraint: The exploitation depth shall not exceed the elevation of -500 meters. 2. Production constraints: Shovel loader capacity constraint: Within each time period (e.g., one year), the total amount of ore mined by the shovel loader shall not exceed its maximum processing capacity. Assume there is a Type A shovel loader in the mine, with a maximum annual processing capacity of 1 million tons. Transportation truck capacity constraint: Within each time period (e.g., one year), the total amount of ore transported by the transportation trucks shall not exceed their maximum transportation capacity. Assume there are two Type B transportation trucks in the mine, each with a maximum annual transportation capacity of 0.5 million tons. 3. Geological constraints: Safety distance constraint: The distance from each exploitation unit to the fault shall not be less than 50 meters. Finally, a classification constraint set is obtained, including a geometric constraint set, a production constraint set, and a geological constraint set.

[0173] Convert the classification constraint set obtained in step S331 into mathematical expressions. The exploitation unit is represented by i, where i = 1, 2,..., 50.

[0174] Geometric constraints: Maximum slope angle constraint: For any two vertically adjacent exploitation units i and j (assuming i is above and j is below), let the elevation difference between their central points be Δh_ij and the horizontal distance be Δd_ij. Then the constraint expression is: arctan(Δh_ij / Δd_ij) ≤ 45°. Minimum exploitation width constraint: Let the width of exploitation unit i be w_i. Then the constraint expression is: w_i ≥ 20 meters. Maximum exploitation depth constraint: Let the elevation of the central point of exploitation unit i be z_i. Then the constraint expression is: z_i ≥ -500 meters.

[0175] Production constraints: Shovel loader capacity constraint: Let the set of exploitation units mined in the t-th year be S_t, and the amount of ore in exploitation unit i be q_i (unit: ton). Then the constraint expression is: Σq_i ≤ 1 million tons (i ∈ S_t). Transportation truck capacity constraint: Let the set of exploitation units transported in the t-th year be S_t, and the amount of ore in exploitation unit i be q_i (unit: ton). Then the constraint expression is: Σq_i ≤ 2 × 0.5 million tons = 1 million tons (i ∈ S_t).

[0176] Geological constraints: Safety distance constraint: Let the distance from exploitation unit i to the fault be d_i. Then the constraint expression is: d_i ≥ 50 meters.

[0177] Finally, a constraint expression is obtained, which is the mathematical expression including all constraint conditions.

[0178] For the constraint expression obtained in step S332, use the exterior point method to construct a penalty function.

[0179] 1. Transformation of inequality constraints into standard form: Transform all inequality constraints into the form of \(g(x)\leq0\).

[0180] Maximum slope angle constraint: \(\arctan(\Delta h_{ij} / \Delta d_{ij}) - 45^{\circ}\leq0\). Minimum mining width constraint: \(20 - w_i\leq0\). Maximum mining depth constraint: \(-500 - z_i\leq0\). Shovel loader capacity constraint: \(\sum q_i - 1000000\leq0\) (\(i\in S_t\)). Haul truck capacity constraint: \(\sum q_i - 1000000\leq0\) (\(i\in S_t\)). Safety distance constraint: \(50 - d_i\leq0\).

[0181] 2. Selection of penalty function: For each inequality constraint \(g(x)\leq0\), select the penalty function \(P(x)=\alpha\times\max(0, g(x))^2\), where \(\alpha\) is the penalty factor.

[0182] 3. Determination of penalty factor: The value of the penalty factor \(\alpha\) needs to be large enough to ensure sufficient penalty for solutions that violate the constraints. Based on experience and tests, the penalty factor \(\alpha\) for all constraints is uniformly set to 1000.

[0183] 4. Construction of penalty function: Penalty function for maximum slope angle constraint:

[0184] \(P1 = 1000\times\max(0, \arctan(\Delta h_{ij} / \Delta d_{ij}) - 45^{\circ})^2\). Penalty function for minimum mining width constraint: \(P2 = 1000\times\max(0, 20 - w_i)^2\). Penalty function for maximum mining depth constraint: \(P3 = 1000\times\max(0, -500 - z_i)^2\). Penalty function for shovel loader capacity constraint: \(P4 = 1000\times\max(0, \sum q_i - 1000000)^2\) (\(i\in S_t\)). Penalty function for haul truck capacity constraint: \(P5 = 1000\times\max(0, \sum q_i - 1000000)^2\) (\(i\in S_t\)). Penalty function for safety distance constraint: \(P6 = 1000\times\max(0, 50 - d_i)^2\). Finally, the penalty function is obtained, including \(P1\), \(P2\), \(P3\), \(P4\), \(P5\), \(P6\).

[0185] Combine the penalty function combinations obtained in step S333 to generate a constraint function. The constraint function consists of a total penalty function, which is the sum of all individual constraint penalty functions. For a given mining sequence, calculate the penalty function values corresponding to all constraint violations in the sequence. For example, if in a mining sequence, two mining units violate the maximum slope angle constraint, three mining units violate the minimum mining width constraint, and the mining volume in the first year exceeds the shovel loader capacity constraint, then the total penalty function value is: Penalty = 2×P1 + 3×P2 + P4; where P1, P2, and P4 are the penalty function values for the corresponding constraints defined in step S333 respectively. The constraint function contains only one function, that is, the total penalty function Penalty. This function comprehensively considers all constraint violation situations and represents the degree of constraint violation of the mining sequence through a scalar value. In the optimization algorithm, add this penalty function value to the objective function value, thereby transforming the constrained optimization problem into an unconstrained optimization problem. Finally, obtain the constraint function {Penalty}. This set contains only one total penalty function.

[0186] Preferably, step S4 includes the following steps:

[0187] Step S41: Construct a simulation environment based on the target mining plan, semantic geological units, and mining units to obtain a mine mining simulation environment;

[0188] Step S42: Use the mine mining simulation environment to simulate the mining process of the target mining plan to obtain mining process simulation data; collect and record the mining process simulation data to obtain the original simulation data record;

[0189] Step S43: Process and integrate the original simulation data record to obtain the processed simulation data; generate a mining performance data set from the processed simulation data to obtain a mining performance data set;

[0190] Step S44: Obtain the actual mining data; perform mining difference analysis on the actual mining data and the mining performance data set to obtain a difference analysis report; determine the geological model parameter correction target according to the difference analysis report to obtain a model correction target set;

[0191] Step S45: Select a correction method according to the model correction target set to obtain a model correction method set; use the model correction target set and the model correction method set to adjust the model parameters of the semantic geological units to obtain an adjusted semantic unit set;

[0192] Step S46: Perform model verification and evaluation according to the adjusted semantic unit set to obtain a model verification report; generate a corrected geological model according to the model verification report and the adjusted semantic unit set.

[0193] In the embodiments of the present invention, a mine mining simulation environment is constructed based on a target mining plan, semantic geological units (SGUSs), and mining units (MUSs). An ore body model is established using 3D mine simulation software. Each semantic unit in the SGUS is imported into the simulation software and its corresponding attributes, such as lithology, grade, density, etc., are assigned. The mining unit division plan defined in the target mining plan is imported into the simulation software, and the ore body model is subdivided into the mining units defined in the MUS. Existing mining equipment in the mine, such as 3 loaders with a bucket capacity of 10 m³ and 5 transport trucks with a load capacity of 50 tons, is added to the simulation environment, and their operating parameters, such as loading speed, transport speed, fuel consumption, etc., are set. The boundary conditions of the simulation environment, such as terrain, groundwater level, climate, etc., are set. In the mine mining simulation environment constructed in step S41, the mining process is simulated according to the mining sequence and parameters specified in the target mining plan. The operating trajectories, loading and unloading operations, and the advancement process of the mining face of each loader and transport truck are simulated. During the simulation process, the position, status (such as working, idle, under repair), ore quantity loaded, ore quantity transported, fuel consumption, etc. of each piece of equipment are recorded every 1 hour. These data are stored as raw simulation data records, for example, saved in the CSV file format. The raw simulation data records obtained in step S42 are processed and integrated. For example, the ore loading quantities of each loader within a month are accumulated to calculate the monthly output of each loader. The ore transport quantities and fuel consumption of each transport truck within a month are calculated to evaluate the transport efficiency. According to the processed simulation data, a mining performance dataset is generated. The mining performance dataset includes the following indicators: monthly ore output, monthly stripping volume, average grade, average mining cost, equipment utilization rate, fuel consumption, etc. The actual mining data of the mine, such as monthly ore output, monthly stripping volume, average grade, average mining cost, etc., are obtained. The actual mining data is compared and analyzed with the mining performance dataset generated in step S43 to calculate the difference between the two. For example, the difference between the actual monthly output and the simulated monthly output is compared, and the reasons for the difference are analyzed. According to the difference analysis report, the geological model parameters that need to be corrected are determined. For example, if the average grade of the actual mining is lower than the simulated result, the estimated value of the grade in the geological model needs to be corrected. The parameters that need to be corrected and their target values are recorded in the model correction target set. For example, "the average grade of ore body A" is added to the model correction target set, and its target value is set to the average grade of the actual mining. According to the model correction target set obtained in step S44, a suitable model correction method is selected. For example, if the model correction target is to increase the average grade of ore body A, the Kriging interpolation method can be used to re-interpolate the grade model of ore body A using the actual mining data. The relevant parameters in the semantic geological unit (SGUS) are adjusted using the selected model correction method.For example, using Kriging interpolation method, recalculate the grade values of each semantic unit in ore body A based on the actual mining data and the variogram model. Verify and evaluate the adjusted semantic unit set in step S45. Import the adjusted semantic unit set into the mine mining simulation environment constructed in step S41, re - conduct the mining process simulation, and generate a new mining performance data set. Compare the new mining performance data set with the actual mining data to evaluate the effect of model correction. Record the verification results in the model verification report. Generate a corrected geological model based on the model verification report and the adjusted semantic unit set. The corrected geological model includes the updated semantic unit set and other relevant geological information, such as faults, lithology distribution, etc.

[0194] Therefore, from any perspective, the embodiments should be regarded as exemplary and non - restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Thus, all changes falling within the meaning and scope of the equivalent elements of the application document are intended to be embraced within the present invention.

[0195] The above - described are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. A method for managing reserves of geological resources in a mine, characterized in that: The following steps are involved: Step S1: Collect multi-source data of mine geology, and perform multi-source data preprocessing to obtain fused geological data; extract geological semantic features from the fused geological data, and perform semantic reconstruction to obtain semantic geological units; Step S2 is specifically as follows: Step S21: determining key geological attributes of the semantic geological unit; Step S22: Calculate the reserve probability of the semantic geological unit based on the Bayesian network structure according to the variation function to obtain unit probability data, wherein step S22 is specifically as follows: Step S221: define the attribute of the semantic geological unit - the reserve node; define the hidden node according to the fused geological data; Step S222: determining the Bayesian network structure of the attribute-reserve nodes and hidden nodes to obtain a Bayesian network structure diagram; Step S223: Acquire historical geological data; determine the prior probability distribution of the Bayesian network structure diagram according to the historical geological data to obtain the node prior probability; Step S224: determining the spatial correlation of geological attributes according to the variation function, and constructing a likelihood function according to the Bayesian network structure diagram and the semantic geological unit to obtain a conditional probability table; Step S225: Perform Bayesian inference on the Bayesian network structure diagram according to the node prior probability and the conditional probability table to obtain the node posterior probability; Step S226: extracting the posterior probability distribution of the reserve node from the node posterior probability, and generating a unit probability model according to the Bayesian network structure diagram to obtain unit probability data; Step S23: Perform Bayesian inference based on the unit probability data and update the parameters to obtain updated probability data, wherein step S23 is specifically as follows: Step S231: acquiring newly added exploration data; integrating observation data of the newly added exploration data to obtain integrated observation data; Step S232: performing Gibbs sampling on the Bayesian network structure diagram according to the conditional probability table and the integrated observation data to obtain a posterior sample; Step S233: Calculate the posterior distribution according to the posterior samples to obtain the node posterior probability; Step S234: updating the graph parameters of the Bayesian network structure graph using the node posterior probability to obtain updated probability data; Step S24: generating reserve probability distribution for semantic geological units according to the updated probability data to obtain unit reserve distribution; Step S25: generating unit reserve probability for unit reserve distribution to obtain unit reserve probability; Step S3 is specifically as follows: Step S31: obtaining mining technical parameters and mining constraints; constructing mining units for semantic geological units according to the mining technical parameters to obtain mining units; Step S32: constructing a mining optimization objective function according to the mining unit and the unit reserve probability to obtain a comprehensive objective function, wherein step S32 is specifically as follows: Step S321: defining a single objective function according to the mining unit and the unit reserve probability to obtain a single objective function, wherein the single objective function includes a reserve maximization objective, a grade balance objective, a geological structure stability objective, and a mining efficiency objective; Step S322: performing single objective function standardization on the single objective function to obtain a standardized single objective function; Step S323: determining the target weight based on fuzzy comprehensive evaluation according to the standardized single objective function to obtain a target weight vector; Step S324: constructing a comprehensive objective function according to the standardized single objective function and the objective weight vector to obtain a comprehensive objective function; Step S33: converting the mining constraint conditions into mathematical expressions, and constructing constraint functions according to the mining units and mining technical parameters to obtain constraint functions; Step S34: taking the comprehensive objective function and the constraint function as inputs of the optimization algorithm, performing multi-objective optimization on the mining unit based on the multi-objective optimization algorithm, and obtaining a mining sequence solution set; Step S35: generating a mining plan for the mining sequence solution set to obtain a mining plan; Step S36: Evaluate the mining plan and perform Pareto optimal solution analysis to obtain a target mining plan; Step S4: simulating the mining process according to the target mining plan to obtain a mining performance data set; performing mining difference analysis on the mining performance data set to obtain a difference analysis report; The geological model is revised according to the difference analysis report and semantic geological units to obtain a revised geological model.

2. The method for managing reserves of geological resources in mines according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: collecting multi-source data of mine geology; performing multi-source data preprocessing on the multi-source data of mine geology to obtain fused geological data; Step S12: constructing an initial three-dimensional geological model for the fused geological data, and performing preliminary grid division to obtain an initial geological grid model; Step S13: extracting geological semantic features from the fused geological data according to the initial geological grid model to obtain a geological feature vector set; Step S14: performing geological semantic segmentation on the initial geological grid model according to the geological feature vector set to obtain an initial semantic unit set; Step S15: Optimize the initial semantic unit set into semantic geological units to obtain semantic geological units.

3. The method for managing reserves of geological resources in mines according to claim 1, characterized in that: Step S224 is specifically as follows: Extract the geological attributes of semantic geological units, calculate the correlation coefficients of geological attributes between different spatial positions according to the variogram, and obtain the spatial correlation of attributes; Kriging estimation is performed based on attribute spatial correlation and semantic geological units to obtain attribute node estimation values; According to the estimated values ​​of the attribute nodes, the attribute node conditional probability table is constructed for the Bayesian network structure graph to obtain the attribute node conditional probability table; The conditional probability table of the attribute node is used to derive the conditional probability distribution of the reserve node and obtain the conditional probability table.

4. The method for managing reserves of geological resources in mines according to claim 1, characterized in that: Step S33 includes the following steps: Step S331: classifying the mining constraints to obtain a classified constraint set, wherein the classified constraint set includes a geometric constraint set, a production constraint set, and a geological constraint set; Step S332: performing constraint mathematical expression on the classification constraint set to obtain a constraint expression; Step S333: selecting an exterior point penalty function for the constraint expression and determining a penalty factor to obtain a penalty function; Step S334: Generate a constraint function for the penalty function to obtain a constraint function.

5. The method for managing reserves of geological resources in mines according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: constructing a simulation environment according to the target mining plan, semantic geological units and mining units to obtain a mining simulation environment; Step S42: using the mining simulation environment to simulate the mining process of the target mining plan to obtain mining process simulation data; collecting and recording the mining process simulation data to obtain original simulation data records; Step S43: Processing and integrating the original simulation data records to obtain processed simulation data; generating a mining performance data set for the processed simulation data to obtain a mining performance data set; Step S44: obtaining actual mining data; performing mining difference analysis on the actual mining data and the mining performance data set to obtain a difference analysis report; determining the geological model parameter correction target according to the difference analysis report to obtain a model correction target set; Step S45: selecting a correction method according to the model correction target set to obtain a model correction method set; adjusting model parameters of the semantic geological unit using the model correction target set and the model correction method set to obtain an adjusted semantic unit set; Step S46: Perform model verification and evaluation based on the adjusted semantic unit set to obtain a model verification report; generate a revised geological model based on the model verification report and the adjusted semantic unit set to obtain a revised geological model.

Images