A method and system for simulating mining areas

Abstract

This invention discloses a method and system for simulating mining areas, relating to the field of ore mining technology. Based on the ore distribution within a theoretical target mining area, an initial set of sampling points is set. Based on this initial set, the ore distribution within each sampling point is obtained. Based on the ore distribution, areas where data fluctuations exceed preset values ​​are identified as key variation areas. This invention provides a solid foundation for efficient and safe mining by progressively refining the theoretical target mining area and determining the actual mining area.

Description

Technical Field

[0001] This invention belongs to the field of ore mining technology, specifically a method and system for simulating mining areas. Background Technology

[0002] With the development of the times and the advancement of technology, computer technology has been widely applied in the field of ore mining, enabling workers to understand the mine conditions before formal mining operations begin, thereby improving mining efficiency. Currently, most technologies employ processes such as ore breaking, loading, transportation, face support, and roof control to complete ore mining, but these methods are relatively inefficient.

[0003] In existing technologies, when simulating mining areas, the common practice is to pre-select sampling points for simulation. However, in actual simulations, the initial sampling points often fail to accurately match real-world requirements, leading to discrepancies between the simulation results and reality. Therefore, effectively correcting inappropriate sampling points has become a pressing problem that needs to be solved. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for simulating mining areas, which solves the technical problem that existing technologies cannot effectively correct sampling points.

[0005] A method for simulating mining areas includes:

[0006] S1. Based on the ore distribution within the theoretical target mining area, set an initial sampling point set {P1, P2, P3, P4, ..., P...}. N}, where N is a positive integer greater than or equal to 1;

[0007] S2. Based on the initial sampling point set, obtain the ore distribution at each sampling point;

[0008] S3. Based on the ore distribution, areas where data fluctuations exceed preset values ​​are identified as key variation areas.

[0009] S4. Set up several new sampling points in the key variation area, and repeat S2-S3 until the fluctuation of ore distribution data no longer exceeds the preset value, and generate a candidate sampling point set.

[0010] S5. Evaluate the candidate sampling point set to determine the final sampling point set;

[0011] S6. Generate the actual mining area based on the final sampling point set.

[0012] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0013] S41. Construct a spatial distribution database of ore bodies based on existing geological data, including three-dimensional geological models, historical exploration data, and real-time monitoring information.

[0014] S42. Calculate the gradient of ore grade variation and mark the area where the gradient value exceeds the preset threshold as a layered variation area.

[0015] S43. Within the layered variation region, locate the spatial intersection point and the grade extreme point. The spatial intersection point includes the three-dimensional intersection position of different mineral layers and the intersection of mineral layers and structural belts. The grade extreme point includes the maximum grade point and the minimum grade point.

[0016] S44. Prioritize deploying several sampling points at the spatial intersections and grade extreme points to form a sampling point array with non-uniform density;

[0017] S45. Evaluate the contribution of the sampling matrix to reducing the variance of resource quantity estimation. Terminate the sampling optimization process when the contribution meets the preset standard.

[0018] As a further aspect of the present invention, the specific steps of S45 are as follows:

[0019] The system performs conditional simulations on known sampled data to generate multiple equally probable taste distributions.

[0020] Calculate the resource quantity and variance for each implementation;

[0021] Analyze the contribution of newly added sampling points to reducing variance.

[0022] As a further aspect of the present invention: between steps S3 and S4, the following is also included:

[0023] Extract geological model data of similar geological regions from historical records. The geological model data includes lithological assemblage, mineralization patterns, and structural features.

[0024] Establish a geological model matching index system, the index of which includes lithological similarity, structural strike angle, and mineralization element correlation;

[0025] Identify and match geological models obtained from known areas with existing data from key areas;

[0026] When the matching degree exceeds the preset threshold, the geological model of the known area is migrated to the key variation area to supplement the geological information of the sparse data area.

[0027] As a further aspect of the present invention, it also includes:

[0028] Identify layered variation zones where ore layer changes exceed a threshold and structural intersections where ore layers intersect with faults, and mark these two types of areas as key areas;

[0029] Within the aforementioned key areas, based on the curvature of the ore layers and the fault-guided ore potential score, a sampling density M times higher than that of ordinary areas is set, where M times is a positive integer greater than or equal to 2.

[0030] Calculate the impact of the current sampled data on the resource estimation error, and reset the sampling at the position where the error exceeds the first threshold until the error is less than the second threshold.

[0031] As a further aspect of the present invention, it also includes:

[0032] After identifying the layered variation region, the intersection points of the ore layer and the fault in three-dimensional space are marked, and the area with a radius of 20 meters centered on the intersection point is designated as the structural intersection area.

[0033] As a further aspect of the present invention: the method for evaluating the candidate sampling point set in S5 is as follows:

[0034] Arrange the completed sampling points in chronological order, analyze the location, ore grade, and surrounding geological conditions of each sampling point, determine whether there is an influence relationship between adjacent sampling points, mark the mutually influencing sampling points with lines, and form a visual influence relationship network diagram.

[0035] In the influence relationship network diagram, identify sampling points whose ore grade is higher than the surrounding area and have a strong influence relationship with multiple surrounding sampling points, but whose own influence range is narrow. Mark these points and their surrounding preset range as information cocoon areas.

[0036] For candidate sampling points located within information cocoon areas, their sampling priority is reduced; for candidate sampling points far from all information cocoon areas, their sampling priority is increased.

[0037] Using intelligent planning algorithms, the sampling order that can obtain the most effective quality information is planned by comprehensively considering the priority of candidate points, sampling cost, and distance from already sampled points.

[0038] After completing a new sampling, the influence relationship between the sampled points is re-analyzed, the identification results of the information cocoon area are updated, the priority of the remaining candidate points is adjusted synchronously, and the subsequent sampling path is continuously optimized.

[0039] As a further aspect of the present invention, the method for determining the influence relationship between sampling points is as follows: if high-quality data from a sampling point causes multiple subsequent sampling points to cluster in its vicinity, then the two are considered to have a strong influence relationship; if the data from a certain sampling point has no obvious guiding effect on the surrounding sampling decisions, then the influence relationship is considered to be weak.

[0040] As a further aspect of the present invention: when planning the sampling path, the intelligent planning algorithm prioritizes candidate sampling points that can break the existing information cocoon distribution and fill the gaps in geological information.

[0041] On the other hand, the present invention also proposes a method for simulating mining areas, applicable to the above-mentioned method for simulating mining areas, the method comprising the following steps.

[0042] The initial setup module sets up an initial sampling point set {P1, P2, P3, P4, ..., P...} based on the ore distribution within the theoretical target mining area. N}, where N is a positive integer greater than or equal to 1;

[0043] The acquisition module obtains the ore distribution at each collection point based on the initial sampling point set.

[0044] The module identifies regions where data fluctuations exceed preset values ​​as key variation areas based on ore distribution.

[0045] The sampling point generation module sets several new sampling points in key variation areas and repeats S2-S3 until the fluctuation of ore distribution data no longer exceeds the preset value, generating a candidate sampling point set.

[0046] The result determination module evaluates the candidate sampling point set and determines the final sampling point set;

[0047] The generation module generates the actual mining area based on the final sample point set.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] This invention constructs an overall understanding from initial sampling, identifies key variation areas based on data fluctuations, and then refines the sampling through iterative sampling to gradually and accurately locate areas with significant variations in ore grade, potential high-value ore bodies, or complex geological conditions. This helps avoid blind exploration, improves the targeting and accuracy of resource exploration, and by dynamically adjusting the sampling strategy, it intensifies sampling in key areas to ensure that no valuable mineralization information is overlooked, while reducing invalid sampling in ordinary areas. This allows exploration data to more accurately and comprehensively reflect ore distribution. On the one hand, the non-uniform sampling strategy avoids wasting exploration resources in unnecessary areas and reduces exploration costs; on the other hand, by comprehensively evaluating the economics, representativeness, and mining feasibility of sampling points, high-cost, low-value sampling points are eliminated, further reducing potential cost waste in subsequent mining stages. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the method framework structure of the present invention. Detailed Implementation

[0051] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Please see Figure 1 This application provides a method for simulating mining areas, including:

[0053] S1. Based on the ore distribution within the theoretical target mining area, set an initial sampling point set {P1, P2, P3, P4, ..., P...}. N}, where N is a positive integer greater than or equal to 1;

[0054] S2. Based on the initial sampling point set, obtain the ore distribution at each sampling point;

[0055] S3. Based on the ore distribution, areas where data fluctuations exceed preset values ​​are identified as key variation areas.

[0056] S4. Set up several new sampling points in the key variation area, and repeat S2-S3 until the fluctuation of ore distribution data no longer exceeds the preset value, and generate a candidate sampling point set.

[0057] S5. Evaluate the candidate sampling point set to determine the final sampling point set;

[0058] S6. Generate the actual mining area based on the final sampling point set.

[0059] Specifically, in the early planning stages of mining operations, it is generally necessary to first gain a preliminary understanding of the target mining area. Typically, geological exploration personnel will construct a rough model of the ore distribution within the theoretical target mining area based on geological theories, empirical data from surrounding mines, geophysical exploration (such as gravity exploration and magnetic exploration), and geochemical measurements (such as soil and stream sediment sampling and analysis). This model includes predictions about ore types, possible enrichment areas, and the correlation between geological structures and ore distribution.

[0060] Furthermore, when setting up sampling points, it is important to ensure that the sampling points cover the entire target mining area as much as possible to guarantee a relatively complete understanding of the ore distribution within the area. A regular grid method can be used, where sampling points are evenly distributed within the mining area at certain intervals (e.g., every 50 meters × 50 meters); alternatively, sampling points can be appropriately densified in areas with potential ore enrichment, such as faults and folds, based on geological structural characteristics.

[0061] Furthermore, in practical applications, considering the difficulty of actual exploration operations, the location of sampling points needs to be easily accessible to exploration personnel, avoiding areas with overly rugged terrain or difficult access. The initial number of sampling points is represented by N, where N is a positive integer greater than or equal to 1, and the specific value depends on factors such as the size and complexity of the mining area and the exploration budget.

[0062] Furthermore, after setting up the initial sampling points, the actual data collection phase begins. Exploration personnel will carry specialized sampling equipment to each sampling point to carry out the work.

[0063] Exploration personnel select appropriate sampling methods based on the type of mine. For open-pit mines, methods such as borehole sampling and groove sampling may be used to collect ore samples from different depths downwards from the surface. For underground mines, sampling is required on the rock walls within the tunnels. The number and depth of samples collected at each sampling point are determined based on specific circumstances. Generally, multiple samples at different depths are collected to obtain information on the vertical distribution of ore at that point.

[0064] The collected ore samples are sent to the laboratory where chemical analysis (such as X-ray fluorescence spectroscopy and atomic absorption spectroscopy) and physical analysis (such as density measurement and hardness testing) are used to determine the grade (content of useful minerals), mineral composition, and structure of the ore in the samples. By analyzing the samples from each sampling point, the distribution of ore at each sampling point can be obtained, and this data will serve as the basis for subsequent analyses.

[0065] After obtaining the ore distribution data from each sampling point, the data needs to be analyzed and processed to identify areas where there are significant changes in ore distribution. Here, the degree of change in ore distribution is measured by calculating data fluctuations. Statistical methods can be used, such as calculating the difference in ore grade between adjacent sampling points, the standard deviation, and other indicators. For example, the average difference in ore grade between each sampling point and sampling points within a certain range (such as four adjacent sampling points) can be calculated as the data fluctuation value for that point.

[0066] A preset value is a standard for measuring whether data fluctuations are significant, and its setting needs to be combined with the actual situation and experience of the mine. For mines where ore grade changes are relatively stable, the preset value can be set relatively low; while for mines where ore grade changes significantly, the preset value can be appropriately increased.

[0067] When the data fluctuation values ​​of multiple sampling points within a certain area exceed a preset value, that area is identified as a critical variation area. These areas indicate significant uncertainty in ore distribution, potentially concealing high-grade ore enrichment zones or presenting complex geological conditions and mining difficulties, thus requiring further in-depth exploration.

[0068] Furthermore, for the identified key variation areas, in order to more accurately grasp the distribution of ore, it is necessary to add new sampling points in these areas. The arrangement of new sampling points also needs to take into account the regional characteristics. A denser grid arrangement can be adopted, or the sampling points with large data fluctuations can be densely arranged around them.

[0069] After adding a new sampling point, steps S2 and S3 are repeated, namely, collecting and analyzing ore samples from the new sampling point to calculate the new data fluctuation value. This process is repeated until the data fluctuation value of all sampling points within the key variation area does not exceed the preset value. At this point, the set of all sampling points, including the initial sampling point and the newly added sampling point, constitutes the candidate sampling point set. The sampling point data in this set can accurately and comprehensively reflect the ore distribution within the target mining area.

[0070] Although the candidate sampling point set can reflect the ore distribution well, in actual mining, it is still necessary to comprehensively consider other factors to evaluate and screen it in order to determine the final sampling point set;

[0071] Finally, based on the distribution of ore grades in the sampling data and the economic boundary grade (i.e., the minimum grade at which mining can be profitable), the actual mining boundary is determined. Areas where the ore grade is higher than the economic boundary grade are designated as the actual mining area.

[0072] Furthermore, within the defined mining area, based on the geological conditions (such as rock hardness and stability) reflected by sampling points, mining methods (such as open-pit mining and underground mining), mining sequence, and mining technology are planned. Through these steps, the theoretical target mining area is gradually refined and the actual mining area is determined, providing a solid foundation for efficient and safe mining.

[0073] As an optional embodiment, the specific steps of S4 are as follows:

[0074] S41. Construct a spatial distribution database of ore bodies based on existing geological data, including three-dimensional geological models, historical exploration data, and real-time monitoring information.

[0075] It should be noted that the three-dimensional geological model is constructed by using geological exploration software (such as Surpac and 3DMine) based on borehole data, seismic exploration and other data to create a three-dimensional visualization model that includes stratigraphic structure, rock type and tectonic zone distribution, and intuitively presents the spatial morphology and geological environment of the ore body.

[0076] Historical exploration data, through the collection of past sampling and analysis results and geological survey reports, covers information such as ore grade, mineral composition, and lithological changes, providing historical references for the analysis of ore body characteristics;

[0077] Real-time monitoring information is obtained by collecting mine environmental data in real time through sensor networks (such as borehole stress sensors and groundwater level monitors), which dynamically reflects the current state of the ore body and ensures the timeliness of the database information.

[0078] Furthermore, building a database facilitates providing unified data support for subsequent steps. For example, when calculating the ore grade variation gradient in S42, the spatial coordinates in the three-dimensional model and the grade values ​​in historical data can be directly retrieved, enabling efficient use and cross-validation of the data.

[0079] S42. Calculate the gradient of ore grade variation and mark the area where the gradient value exceeds the preset threshold as a layered variation area.

[0080] S43. Within the layered variation region, locate the spatial intersection point and the grade extreme point. The spatial intersection point includes the three-dimensional intersection position of different mineral layers and the intersection of mineral layers and structural belts. The grade extreme point includes the maximum grade point and the minimum grade point.

[0081] Among them, the intersection of different mineral layers is the spatial contact zone of different mineral layers identified by the three-dimensional geological model. These locations often form rich ore sections due to the superposition of mineralization processes.

[0082] Intersection of ore layers and structural zones: Structural zones (such as faults and fold axes) are channels for the migration and sedimentation of ore-forming fluids. The ore grade changes drastically at the intersection of these zones and ore layers, which are potential high-grade enrichment areas.

[0083] The extreme value points of the grade are determined by searching for local maximum grade points (high-grade core area) and minimum grade points (grade depletion area) based on the interpolated grade distribution surface. These points can reveal extreme changes in the grade of the ore body and are crucial for resource estimation.

[0084] S44. Prioritize deploying several sampling points at the spatial intersections and grade extreme points to form a sampling point array with non-uniform density;

[0085] S45. Evaluate the contribution of the sampling matrix to reducing the variance of resource quantity estimation. Terminate the sampling optimization process when the contribution meets the preset standard.

[0086] Specifically, spatial interpolation algorithms (such as Kriging interpolation) are used to expand the discrete sample point grade data into a continuous spatial distribution surface. Based on this, partial derivatives are used to calculate the rate of grade change at each point in three-dimensional space (X, Y, Z directions), and a gradient vector of ore grade change is synthesized, with its magnitude representing the degree of drastic change.

[0087] A preset threshold is used to identify a layered variation area when the gradient value in a certain region is generally higher than the preset threshold. These areas usually correspond to the boundaries of ore layers, tectonic activity zones, or overlapping parts of different mineralization stages, where ore grades vary complexly and require focused exploration.

[0088] Furthermore, based on the varying importance of geological feature points, a differentiated sampling strategy is implemented to further identify key geological feature points. Sampling points are prioritized at spatial intersections and extreme grade points to ensure exploration accuracy in critical areas. For other areas, the sampling density is appropriately reduced, forming a non-uniform sampling array characterized by "high density in key areas and sparseness in general areas." For example, a sampling point is placed every 10 meters at the intersection of the ore layer and the fault, while a point is placed every 50 meters in stable ore layer areas. By accurately locating key points, sampling efficiency is improved, which helps avoid wasting exploration resources in non-critical areas and provides core data support for subsequent resource estimation. Compared to random placement of points throughout the variable area, this strategy can more efficiently capture grade abrupt changes and extreme values ​​in the ore layer contact zone, thus improving the accuracy of subsequent mining planning.

[0089] It should be understood that by calculating the contribution rate and setting a contribution rate threshold, when this standard is reached, the current sampling point set is considered to meet the accuracy requirements for resource estimation, and the sampling optimization process is terminated. The resulting sampling point set will then serve as a candidate sampling point set and enter the comprehensive evaluation stage of S5. This effectively solves the problems of large sampling blindness, high resource estimation error, waste of exploration costs, and insufficient basis for mining decisions that exist in traditional mine exploration.

[0090] As an optional embodiment, the specific steps of S45 are as follows:

[0091] The system performs conditional simulations on known sampled data to generate multiple equally probable taste distributions.

[0092] Calculate the resource quantity and variance for each implementation;

[0093] Analyze the contribution of newly added sampling points to reducing variance.

[0094] Furthermore, the conditional simulation is based on known sampling data (including ore grade and location information of initial sampling points and newly added sampling points), and uses geostatistical methods to construct a stochastic model that conforms to the actual geological characteristics, thereby generating multiple equally probable ore grade distributions.

[0095] Based on the Kriging interpolation method, and considering the spatial variability of ore grade (described by a variogram), multiple possible grade distribution scenarios are simulated under the premise of satisfying known sampling data constraints. Each implementation represents a grade distribution state of the ore body that conforms to current exploration knowledge.

[0096] Using professional geological modeling software (such as GSLIB and GeostatisticalAnalyst), the coordinates of known sampling points, grade values, and variogram parameters are input. Through Monte Carlo simulation technology, 100-500 sets of grade distributions with equal probability are randomly generated to cover the uncertainty range of ore body grade distribution.

[0097] Furthermore, based on the orebody boundaries defined by the three-dimensional geological model and combined with the grade data from various grade distributions, resource estimation methods such as the block method and the distance power inverse ratio method are used to calculate the ore resource quantity (unit: tons or cubic meters) for each realized ore body. For example, the orebody is divided into multiple regular blocks, and the resource quantity is estimated and accumulated based on the grade and volume of each block.

[0098] Statistical analysis was performed on all realized resource quantities, and their variance was calculated. The formula for calculating variance is: Where n is the total number of simulations, x i Let i be the resource quantity for the i-th implementation. This is the average of all realized resource quantities. A larger variance indicates higher uncertainty in the resource quantity estimation results.

[0099] The formula for calculating contribution is as follows: in, and The values ​​represent the variances of resource quantity estimates before and after sampling, respectively. The higher the contribution value, the more significant the effect of adding sampling points on reducing resource quantity estimation errors.

[0100] Furthermore, contribution calculation helps avoid wasting exploration costs by blindly increasing sampling points, while ensuring that the accuracy of resource estimation meets the needs of mining planning. For example, if the contribution rate is only 2% after adding a new sampling point in a certain area, it indicates that the exploration information in that area is already sufficient, and there is no need to continue to increase sampling density. Resources can be invested in other key areas.

[0101] As an optional embodiment, the step between S3 and S4 further includes:

[0102] Extract geological model data of similar geological regions from historical records. The geological model data includes lithological assemblage, mineralization patterns, and structural features.

[0103] Establish a geological model matching index system, the index of which includes lithological similarity, structural strike angle, and mineralization element correlation;

[0104] Identify and match geological models obtained from known areas with existing data from key areas;

[0105] When the matching degree exceeds the preset threshold, the geological model of the known area is migrated to the key variation area to supplement the geological information of the sparse data area.

[0106] It should be understood that integrating historical exploration reports from mining enterprises, regional geological survey data, and geological research results from scientific research institutions to construct a database covering various geological conditions, with a focus on extracting data such as lithological assemblage (e.g., sandstone-shale interbedded layers, granite intrusions), mineralization patterns (e.g., mineralization stratification characteristics of stratabound deposits, alteration zoning of hydrothermal deposits), and structural characteristics (fault strike, fold morphology, joint development), can reflect the typical characteristics of a geological region;

[0107] Based on the structured data obtained by S2 for key variation areas, such as three-dimensional spatial coordinates (X,Y,Z), lithological codes (e.g., using GB / T17412 standard codes), main structural strike values ​​(°), and mineral element content (ppm), unstructured descriptions (e.g., fault properties, alteration types) are processed into feature vectors (e.g., keywords are extracted through natural language processing and mapped to numerical labels) to form feature vectors for algorithm calculation.

[0108] Based on geological expertise, differentiated weights are set for lithological assemblage, structural strike, and mineralization elements. A weighted Euclidean distance is used to calculate the comprehensive similarity score, which is an existing technology and will not be elaborated on here. This allows for the selection of areas with similar geological characteristics from historical databases, thus narrowing the data retrieval scope.

[0109] Furthermore, to quantitatively assess the geological similarity between known areas and key variation areas, a geological model matching index system was constructed:

[0110] Lithological types are coded (e.g., sandstone = 1, shale = 2), and similarity is measured by calculating the proportion of overlap between key variation areas and known lithological assemblages, for example, using the formula: Where A and B represent the lithological assemblage sets of two regions, and the closer the value is to 1, the higher the lithological similarity.

[0111] The strike angle of a structure is measured using principles of structural geology, taking the angle between the strike of a key variation region and the strike of major known structures (such as faults and fold tracks). A smaller angle indicates a greater similarity in the tectonic stress field and tectonic evolution process. For example, when the angle is less than 15°, the structural strikes are considered to be highly similar.

[0112] The distribution characteristics of major mineralizing elements (such as copper, iron, and gold) in ores from two regions were analyzed, and the linear correlation between element contents was calculated using the Pearson correlation coefficient. For example, The Pearson correlation coefficient is derived, where x i y represents the content of a certain mineralization element in the i-th sample within the first region (usually a key variation region or the region under study). iThe same mineralization element content in the i-th sample within the second region (generally a similar geological region extracted from historical records). The mean element content is denoted by , n represents the number of samples involved in the calculation, and r ranges from -1 to 1. The closer the absolute value is to 1, the stronger the linear correlation between the mineralized element content distributions of the two regions.

[0113] Furthermore, existing geological data (such as lithological descriptions of sampling points, structural measurement data, and elemental analysis results) and historical data of known areas in key variation regions are standardized to unify data formats and dimensions, ensuring comparability. The values ​​of indicators such as lithological similarity, structural strike angle, and mineralization element correlation are calculated separately, and weights are assigned according to the importance of each indicator to the geological model judgment (e.g., lithological similarity weight 0.4, structural strike angle weight 0.3, mineralization element correlation weight 0.3). The comprehensive matching degree is obtained by weighted summation, calculated using the following formula: Wherein, ω1, ω2, and ω3 are the weights of each index, and θ is the angle of the construction direction. All of the above are existing technologies and will not be elaborated on here.

[0114] Furthermore, when the overall matching degree exceeds a preset threshold (e.g., 0.7, set according to the actual geological complexity), the geological models of the known area and the key variation area are considered similar, and model migration can be performed: geological model data from the known area (such as lithological predictions for unexplored areas and the location of potential mineralization enrichment zones) are migrated to the key variation area to fill information gaps in sparse data areas. For example, in an unsampled area within the key variation area, the lithology is inferred to be sandstone based on the migrated geological model, and potential mineralization layers are predicted. The supplemented geological information will serve as an important reference for setting new sampling points in the key variation area in step S4. For example, sampling points can be densified near predicted mineralization enrichment zones to improve exploration efficiency and the accuracy of resource estimation.

[0115] As an optional embodiment, it also includes:

[0116] Identify layered variation zones where ore layer changes exceed a threshold and structural intersections where ore layers intersect with faults, and mark these two types of areas as key areas;

[0117] Furthermore, using the orebody spatial distribution database constructed in step S41, parameters such as ore layer thickness and ore grade are extracted. When the variation of these parameters in the vertical or horizontal direction exceeds a set threshold (e.g., ore layer thickness variation rate exceeds 15%, grade fluctuation exceeds 20%), it is identified as the currently selected layered variation zone. These areas typically correspond to areas with abrupt changes in ore layer depositional environment and strong subsequent alteration, and are key locations for ore body enrichment or morphological changes.

[0118] It should be understood that geological model matching (the new step between S3 and S4) and structural feature analysis are used to identify the intersection areas of ore layers and faults. Faults, as channels for the migration of ore-forming fluids and locations for ore bodies, often form high-grade ore bodies at their intersections with ore layers. By using data on the strike of structural zones and the distribution of ore layers in the database, the spatial intersection points of the two are calculated and marked as structural intersection points.

[0119] Within the aforementioned key areas, based on the curvature of the ore layers and the fault-guided ore potential score, a sampling density M times higher than that of ordinary areas is set, where M times is a positive integer greater than or equal to 2.

[0120] Among them, the ordinary area refers to the ore body area that is not marked as a layered variation area or a structural intersection, and the sampling density is the basic grid spacing;

[0121] If the degree of curvature of the ore layer exceeds the preset threshold and / or any parameter value of the fault-guided ore probability score exceeds the threshold, the sampling density is set according to the corresponding M value.

[0122] Among them, the fault-guided mineralization probability score is obtained by assigning values ​​to multiple key factors such as fault size, nature, activity period, filling material characteristics and regional geological background, and then summing them up. All of the above calculation methods can be implemented using existing algorithms, which will not be elaborated on here.

[0123] Furthermore, the curvature of the ore layer is calculated using a three-dimensional geological model (e.g., by fitting the surface of the ore layer with the least squares method and calculating the curvature value). The larger the curvature value, the greater the degree of bending, indicating that the ore layer is more strongly modified by tectonic stress and the ore body morphology is more complex, requiring a higher sampling density.

[0124] Based on geological model matching results, and combined with factors such as fault size (length, displacement), infill material properties (e.g., whether the fault breccia contains mineralization and alteration), and the distribution of known ore bodies in the surrounding area, the probability of fault-induced mineralization is assessed through expert scoring or machine learning models. For example, if the fault is large, the infill material contains sulfide alteration, and there are rich ore bodies in the adjacent area, then the probability of fault-induced mineralization is considered high.

[0125] For critical areas, the sampling density will be increased to M times that of ordinary areas (M≥2). For example, if ordinary areas are sampled with a 50m×50m grid, critical areas will use a 25m×25m or denser grid to ensure that details of ore body changes under complex geological conditions are captured.

[0126] Based on real-time sampling data feedback, if good mineralization continuity is found in a key area, the M value can be appropriately reduced; conversely, if the mineralization changes drastically, the M value can be further increased to achieve dynamic optimization of sampling density.

[0127] Calculate the impact of the current sampled data on the resource estimation error, and reset the sampling at the position where the error exceeds the first threshold until the error is less than the second threshold.

[0128] Furthermore, traditional resource estimation methods such as the block method and kriging method are used, combined with current sampling data, to calculate the ore body resource quantity. For example, the block method divides the ore body into regular blocks, estimates the resource quantity based on the block grade and volume, calculates the relative error between the estimated resource quantity and the actual resource quantity (if some verification data is known), or calculates the standard deviation of the estimation results of different sampling subsets through cross-validation as an error metric. All of the above algorithms are existing mature technologies and will not be elaborated on here.

[0129] Set a high error tolerance value (e.g., 15%). When the current sampling data causes the resource quantity estimation error to exceed this threshold, it indicates that the sampling information is insufficient and resampling is required at the error limit location.

[0130] Set a low target error value (e.g., 5%), and continue iterating the sampling process until the error is less than the threshold. If the current sampled data is considered sufficient to support a reliable resource estimation, the sampling optimization is terminated.

[0131] As an optional embodiment, it also includes:

[0132] After identifying the layered variation region, the intersection points of the ore layer and the fault in three-dimensional space are marked, and the area with a radius of 20 meters centered on the intersection point is designated as the structural intersection area.

[0133] It should be understood that, by employing spatial geometric algorithms for intersection point location, the ore layer is first treated as a three-dimensional surface (generated through Kriging interpolation or least squares fitting), and the fault is abstracted as a line or surface in three-dimensional space (selected according to the fault type, such as modeling a plate fault as a plane and a complex fault as a surface). Intersection point solving algorithms (such as vector algebra or numerical iteration) are then used to calculate the coordinates of the intersection points between the ore layer surface and the fault line / surface. For example, for plate faults, the plane equation and the ore layer surface equation can be solved simultaneously; for complex faults, ray tracing can be used, emitting rays from the fault boundary towards the ore layer and detecting the intersection points of the rays and the ore layer.

[0134] Furthermore, to avoid misjudgments due to data errors, an error tolerance mechanism is introduced. An allowable coordinate error range (e.g., ±0.5 meters) is set, and a valid intersection point is only confirmed when the calculated coordinates remain within the error tolerance range after multiple iterations or verifications using different algorithms.

[0135] Using the effective intersection point as the center, a spherical region with a fixed radius (20 meters) is delineated as the structural intersection point. This radius is chosen by comprehensively considering geological engineering experience and exploration costs; the 20-meter range covers the potential mineralization enrichment area under the influence of the fault while avoiding an excessively large area that would increase exploration costs. For special geological conditions (such as large fault fracture bands or wide-ranging mineralization anomalies), the radius can be dynamically adjusted through expert experience or machine learning models (trained based on historical exploration data).

[0136] When the designated area overlaps with the ore body boundary or other key areas, Boolean operations are used to merge the boundaries. For example, if the structural intersection overlaps with the layered variation area, the union of the two is taken as the new key area, which helps to ensure that no mineralized favorable areas are missed; if the area exceeds the ore body boundary, the area is cut off with the ore body boundary as the limit.

[0137] As an optional embodiment, the method for evaluating the candidate sampling point set in S5 is as follows:

[0138] Arrange the completed sampling points in chronological order, analyze the location, ore grade, and surrounding geological conditions of each sampling point, determine whether there is an influence relationship between adjacent sampling points, mark the mutually influencing sampling points with lines, and form a visual influence relationship network diagram.

[0139] It should be understood that a structured dataset is formed by sorting the samples in chronological order of their completion.

[0140] Spatial analysis and geostatistical methods are used to determine the influence relationships between adjacent sampling points. For example: Spatial distance: A distance threshold is set (e.g., within 30 meters is considered adjacent); the closer the distance, the greater the likelihood of influence. Grade variation: The difference in ore grade between adjacent sampling points is calculated; if the difference exceeds a preset threshold (e.g., grade fluctuation exceeds 15%) and shows a regular change (e.g., increasing / decreasing), an influence relationship is determined. Geological characteristics: If adjacent points have similar lithology and structural types, or are located on the same vein or fault zone, a geological influence connection is identified. Visualization: Graph theory algorithms (e.g., Dijkstra's algorithm) are used to abstract sampling points into nodes and influence relationships into lines. A visual network diagram is generated using professional geological mapping software (e.g., ArcGIS, Grameme) to intuitively display the spatial and geological relationships between sampling points.

[0141] In the influence relationship network diagram, identify sampling points whose ore grade is higher than the surrounding area and have a strong influence relationship with multiple surrounding sampling points, but whose own influence range is narrow. Mark these points and their surrounding preset range as information cocoon areas.

[0142] Within this framework, a pre-defined circular area (radius 15 meters, adjustable based on mineralization continuity) is defined centered on sampling points that meet the aforementioned characteristics. This area is designated as the information cocoon region. These regions often exhibit data deviations from the overall trend due to local geological anomalies (such as small-scale lenses or localized alteration), and oversampling may interfere with the overall resource estimation.

[0143] For candidate sampling points located within information cocoon areas, their sampling priority is reduced; for candidate sampling points far from all information cocoon areas, their sampling priority is increased.

[0144] Specifically, the priority of candidate sampling points within the region is reduced to 50% of the original level to reduce invalid duplicate sampling. For candidate points that are more than 50 meters away from all information cocoon areas, the priority is increased to 130% of the original level to encourage exploration of uncovered areas. Each candidate point is assigned a priority weight (between 0 and 1) and combined with sampling costs (quantitative indicators such as equipment transportation and manpower input) and distance from already sampled points (Euclidean distance) to form multi-dimensional evaluation parameters.

[0145] Using intelligent planning algorithms, the sampling order that can obtain the most effective quality information is planned by comprehensively considering the priority of candidate points, sampling cost, and distance from already sampled points.

[0146] Among them, intelligent planning algorithms employ genetic algorithms or ant colony algorithms, with the objective function of "obtaining the most effective qualitative information" and using priority weights, sampling costs, and spatial distances as constraints. For example, genetic algorithms simulate natural selection by performing crossover and mutation operations on sampling point combinations, iteratively optimizing sampling paths, etc., which are existing technologies and will not be elaborated upon further.

[0147] After completing a new sampling, the influence relationship between the sampled points is re-analyzed, the identification results of the information cocoon area are updated, the priority of the remaining candidate points is adjusted synchronously, and the subsequent sampling path is continuously optimized.

[0148] After each new sampling is completed, the influence relationships between sampling points are recalculated based on the new data, and the information cocoon area is updated. If the original information cocoon area is confirmed by the new data to have general mineralization characteristics, its label is removed and the priority of sampling points in the area is restored; otherwise, if a new area matching the characteristics appears, it is included in the management. This process is continuously repeated until the resource estimation error meets the termination criterion of S45.

[0149] In summary, on the one hand, visualizing the influence relationship network helps to accurately identify high-grade but narrow-range "information cocoon" areas, thus avoiding misleading exploration directions due to local anomalies. On the other hand, by lowering the priority of sampling points within information cocoon areas and raising the priority of blank areas, combined with intelligent algorithms to plan the sampling sequence, invalid duplicate sampling is reduced. After each sampling, the influence relationship and information cocoon identification results are updated in real time, continuously optimizing subsequent paths. This adapts to the actual geological conditions of complex ore body morphology and uneven mineralization, and is especially suitable for complex mines with multiple overlapping mineralization processes. It helps to improve the flexibility and accuracy of the exploration process. By prioritizing the exploration of high-value areas, it provides more comprehensive geological data support for delineating actual mining boundaries and designing mining plans, shortening the exploration-to-mining cycle, reducing mining risks caused by insufficient data, and ensuring efficient and safe mine operation.

[0150] 9. As an optional embodiment, the method for determining the influence relationship between sampling points is as follows: if high-quality data from a sampling point causes multiple subsequent sampling points to cluster in its vicinity, then the two are considered to have a strong influence relationship; if the data from a certain sampling point has no obvious guiding effect on the surrounding sampling decisions, then the influence relationship is considered to be weak.

[0151] Among them, high-grade sites indicate local rich ore target areas, driving exploration personnel to increase sampling density around them, reflecting the strong guiding role of the site in exploration decisions;

[0152] It should be understood that if the ore grade at a certain sampling point (let's call it point A) is significantly higher than the regional average, and the newly added sampling points in subsequent exploration show a trend of clustering near point A, then it is determined that point A has a strong influence relationship with the subsequent sampling points.

[0153] The percentage of newly added sampling points ≤30 meters from point A is counted. A threshold is set (e.g., P≥50%). If the percentage exceeds the threshold, a strong influence relationship is identified, and the points are marked on the network diagram with arrows (the arrows point to the affected sampling points).

[0154] If the grade data of a certain sampling point (let's call it point B) is close to the regional average, and the subsequent newly added sampling points are evenly distributed in space and do not show a trend of clustering towards point B (e.g., the proportion of newly added points within 30 meters of point B is <20%), then it is determined that point B has a weak influence on the surrounding sampling decisions.

[0155] The percentage of newly added sampling points ≤30 meters from point B (denoted as Q) is counted; a threshold is set (e.g., Q≤20%), and points below the threshold are considered to have a weak influence relationship, and the connection is not marked in the network diagram or is represented by a dashed line.

[0156] As an optional embodiment, when planning the sampling path, the intelligent planning algorithm prioritizes candidate sampling points that can break the existing information cocoon distribution and fill the gaps in geological information.

[0157] It should be understood that when planning sampling paths, intelligent planning algorithms are beneficial in prioritizing candidate sampling points that can break the existing information cocoon distribution (such as being far from high-grade clusters or revealing new geological features) and fill gaps in geological information (such as areas with high prediction errors or vein extension directions). By dynamically optimizing the path through multi-dimensional evaluation matrices (indicators such as fusion distance, grade difference, and error reduction) and particle swarm optimization algorithms, the exploration resources can be tilted towards areas with high information value. This helps to avoid local data biases, improve the accuracy of global resource estimation, and reduce exploration costs.

[0158] Secondly, the present invention also proposes a method for simulating mining areas, the method comprising the following steps:

[0159] The initial setup module sets up an initial sampling point set {P1, P2, P3, P4, ..., P...} based on the ore distribution within the theoretical target mining area. N}, where N is a positive integer greater than or equal to 1;

[0160] The acquisition module obtains the ore distribution at each collection point based on the initial sampling point set.

[0161] The module identifies regions where data fluctuations exceed preset values ​​as key variation areas based on ore distribution.

[0162] The sampling point generation module sets several new sampling points in key variation areas and repeats S2-S3 until the fluctuation of ore distribution data no longer exceeds the preset value, generating a candidate sampling point set.

[0163] The result determination module evaluates the candidate sampling point set and determines the final sampling point set;

[0164] The generation module generates the actual mining area based on the final sample point set.

[0165] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of simulating a mining region, characterized by, Comprise: S1, set an initial sample point set based on the ore distribution in the theoretical target mine mining area wherein N is a positive integer greater than or equal to 1. S2, based on the initial sampling point set, the distribution of ore in each collection point is obtained; S3, based on the distribution of ore, the area where the data fluctuation exceeds the preset value is determined as the key variation area; S4, set several new sampling points in the key variation area, and repeat S2-S3 until the ore distribution data fluctuation does not exceed the preset value, and generate a candidate sampling point set; S5, evaluate the candidate sampling point set to determine the final sampling point set; S6, based on the final sampling point set, the actual mining area of the mine is generated; The specific steps of S4 are: S41, based on the existing geological data information, a spatial distribution database of ore body is constructed, wherein the geological data information includes a three-dimensional geological model, historical exploration data and real-time monitoring information; S42, calculate the grade change gradient of the ore, and mark the area where the gradient value exceeds the preset threshold as the layered variation area; S43, in the layered variation area, locate the spatial intersection point and the grade extreme point, the spatial intersection point includes the three-dimensional intersection position of different ore layers, the intersection position of ore layer and structure belt, and the grade extreme point includes the maximum and minimum value points; S44, preferentially deploy several sampling points at the spatial intersection point and the grade extreme point to form a non-uniform density sampling point array; S45, evaluate the contribution of the sampling point array to reducing the variance of resource quantity estimation, and terminate the sampling optimization process when the contribution meets the preset standard.

2. A method of simulating a mining region according to claim 1, characterised in that, The specific steps of S45 are: Conditional simulation is performed on the known sampling data to generate multiple equal-probability grade distribution implementations; Calculate the resource quantity and its variance of each implementation; Analyze the contribution of the newly added sampling points to reducing the variance.

3. A method of simulating a mining region according to claim 1, characterised in that, Between steps S3 and S4, it also includes: Extract the geological pattern data of similar geological areas in the historical records, wherein the geological pattern data includes lithology combination, mineralization rule and structure characteristics; Establish a geological pattern matching index system, which includes lithology similarity, structure strike angle and mineralization element correlation; Identify and match the geological pattern obtained from the known area with the existing data in the key area; When the matching degree exceeds the preset threshold, migrate the geological pattern of the known area to the key variation area to supplement the geological information of the sparse data area.

4. A method of simulating a mining region according to claim 1, characterised in that, It also includes: Identify the layered variation area where the ore layer changes exceed the threshold and the structure intersection position where the ore layer and fault intersect, and mark the two types of areas as key areas; In the above key areas, set the sampling density M times higher than that in ordinary areas according to the bending degree of the ore layer and the fault ore guiding possibility score, wherein M times is a positive integer greater than or equal to 2; Calculate the influence of the current sampling data on the estimation error of resource quantity, and reset the sampling at the position where the error exceeds the first threshold until the error is less than the second threshold.

5. A method of simulating a mining region according to claim 4, characterised in that: It also includes: After determining the layered variation area, mark the intersection point of the ore layer and the fault in the three-dimensional space, and divide the structure intersection position with the intersection point as the center and a radius of 20 meters.

6. A method of simulating a mining region according to claim 1, characterised in that, The evaluation method of the candidate sampling point set in S5 is: The completed sampling points are arranged in chronological order, the position, ore grade, and surrounding geological conditions of each sampling point are analyzed, it is determined whether there is an influence relationship between adjacent sampling points, the sampling points that influence each other are marked with a line, and a visual influence relationship network diagram is formed; In the influence relationship network diagram, identify the sampling points with high ore grade than the surrounding area, and with strong influence relationship with multiple surrounding sampling points, but with narrow influence range, mark these points and their surrounding preset range as information cocoon house area; For candidate sampling points located in the information cocoon house area, reduce their priority sampling level; for candidate points far away from all information cocoon house areas, improve their sampling priority; Using intelligent planning algorithm, considering the priority of candidate points, sampling cost and distance from the sampled points, the sampling sequence that can obtain the most effective geological information is planned; Complete a new sampling, reanalyze the influence relationship between the sampled points, update the identification result of the information cocoon house area, and synchronously adjust the priority of the remaining candidate points, continuously optimize the subsequent sampling path.

7. A method of simulating a mining region according to claim 6, characterised in that, The way to determine the influence relationship between the sampling points is: if the high-grade data of a sampling point leads to the aggregation of multiple subsequent sampling points near it, it is determined that there is a strong influence relationship between them; if the data of a sampling point has no obvious guiding effect on the surrounding sampling decision, it is determined that the influence relationship is weak.

8. A method of simulating a mining region according to claim 6, characterised in that, When planning the sampling path, the intelligent planning algorithm preferentially selects candidate sampling points that can break the existing information cocoon house distribution and fill in the blank area of geological information.

9. A mine exploitation area simulation system suitable for use in a mine exploitation area simulation method according to any one of claims 1 to 8, characterized in that, The system includes the following steps: The initial setting module sets an initial sampling point set based on a distribution of ores in a mining area of a theoretical target mine wherein N is a positive integer greater than or equal to 1. An acquisition module acquires the ore distribution in each collection point based on an initial set of sampling points; A determination module determines a region where data fluctuation exceeds a preset value as a key variation region based on the ore distribution; A sampling point generation module sets a number of new sampling points in the key variation region and repeatedly executes S2-S3 until the ore distribution data fluctuation does not exceed the preset value, generating a candidate sampling point set; A result determination module evaluates the candidate sampling point set to determine a final sampling point set; A generation module generates an actual mining area based on the final sampling point set.

Images