Mineral sample collection optimization method based on stratum change recognition

Abstract

The invention discloses a mineral sample acquisition optimization method based on stratum change identification, and particularly relates to the technical field of mineral exploration sampling, comprising the following steps: acquiring lithologic characteristics, fault distribution, underground stress information and historical sampling data, fusing to form spatial reference data and generating a stratum change level distribution map, then constructing a sample collection candidate area map; generating a sampling path group and adjusting path layout according to the candidate area map, matching a sampling mode for each path, recording multiple types of data in real time in the sampling process and comparing the data with spatial reference data, and dynamically adjusting the sampling path or mode according to a comparison result; spatial reference data are generated through multi-source geological information fusion, and high-precision stratum change recognition is achieved; the sampling path layout is optimized in combination with disturbance distribution and operation constraint, so that the operation safety and the coverage integrity are improved; in the sampling process, response data are compared in real time, a sampling strategy is dynamically adjusted, and the geological consistency and reliability of a sampling result are ensured.

Description

Technical Field

[0001] This invention relates to the field of mineral exploration sampling technology, and more specifically, to an optimized method for mineral sample collection based on stratigraphic change identification. Background Technology

[0002] With the increasing demand for deep resource exploration and the growing complexity of geological conditions, mineral sampling, as a crucial step in the early stages of mineral resource development, directly impacts the scientific and economic viability of subsequent resource evaluation, modeling and prediction, and mining layout. Currently widely used sample collection optimization methods are beginning to incorporate strategies based on stratigraphic change identification. This involves dynamically adjusting sampling paths and locations by identifying information such as the distribution of geological disturbances, change boundaries, and stratigraphic continuity, aiming to improve the representativeness of sampling coverage and the completeness of geological structure representation.

[0003] Current technological practices still have many limitations and shortcomings. In areas with complex or drastically changing stratigraphic structures, existing identification mechanisms often rely on rough spatial similarity or attribute continuity to judge the "repetition" of geological features. When multiple spatial units exhibit apparent consistency such as lithological convergence or waveform similarity, the identification algorithm is prone to misclassifying them as "mined areas" or "non-critical change zones," thus automatically excluding them during path generation or sampling point screening. This "repetition judgment error" will lead to the erroneous omission of abrupt geological change zones or key transition areas, affecting the representativeness of samples and introducing potential data gaps and structural biases in subsequent orebody modeling, resource reserve estimation, and other stages.

[0004] In the process of route planning and dynamic adjustment, the focus is generally on the coverage of the sampling area and the principle of the shortest path, often neglecting the limitations of the sampling equipment's operating conditions and on-site construction safety requirements. Especially in areas with undulating elevations, fault lines, or narrow spaces, if the route construction does not fully consider terrain accessibility, equipment deployment adaptability, and the rationality of spatial arrangement between routes, problems such as overlapping sampling routes and dense deployment conflicts are very likely to occur. Such route interference not only increases the complexity of equipment scheduling and manual intervention, but may also cause equipment collisions, loss of control of construction scheduling, and even serious personnel safety risks. Therefore, this invention proposes a mineral sample collection optimization method based on stratigraphic change identification to solve the above problems. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: An optimized method for mineral sample collection based on stratigraphic change identification includes the following steps: The lithological characteristics, fault distribution, underground stress information and historical sampling data of the target area are acquired and fused to form spatial reference data that reflects the degree of change in stratigraphic structure, and is used to divide regional units of different change levels. Based on spatial reference data, a distribution map of stratigraphic variation levels is generated. By modeling the degree of geological disturbance within regional units, spatial blocks with different degrees of disturbance are marked, and a candidate area map for sample collection is constructed. Based on the candidate area map for sample collection, a sampling path group is generated. During the path generation process, the spatial layout of the path is adjusted based on the distribution of geological disturbance between spatial blocks, combined with the possibility of path intersection and operational safety requirements, so that the path group can meet operational safety conditions while ensuring coverage effectiveness. For each path in the sampling path group, a corresponding sample collection method is matched to make the collection method adaptable to the characteristics of different regions; During the sample collection process, the sampling resistance change value, geological response waveform and preliminary physical properties of the sample at each sampling point are recorded in real time and compared with spatial reference data. The sampling path or sampling method is adjusted according to the comparison results to ensure that the sampling process is consistent with the geological change characteristics.

[0006] In a preferred embodiment, the step of jointly processing lithological characteristic data, fault distribution data, subsurface stress information, and historical sampling data when forming spatial reference data further includes: Multiple geological parameters are converted to the same scale, and the amplitude normalization method makes the data from different sources comparable at the same order of magnitude. The joint change trend between parameters is judged by geological multidimensional correlation analysis. On this basis, the target area is divided into multiple continuous spatial units in a multi-level partitioning method, so that each spatial unit can serve as the basis for subsequent change level determination. When performing multidimensional fusion processing, the direction of geological change is obtained by trend line extraction, and the overall variability within the region is obtained by multi-point joint covariance estimation. This enables the spatial reference data to fully reflect the structural differences in geological change in the region, thereby making the subsequent change level classification more reliable and geologically expressive.

[0007] In a preferred embodiment, the step of modeling the degree of disturbance of each spatial unit when generating a stratigraphic variation level distribution map includes: Information on lithological variation intensity, fault structure characteristics, stress field response trends and historical sampling offsets within spatial units is collected to construct four types of variation factors. The input data for disturbance modeling is calculated through multi-factor joint analysis. During the perturbation modeling process, multi-factor fusion is used for unit-level evaluation, and gradient identification is performed at the spatial unit boundary to extract the geological structure abrupt boundary features. The disturbance assessment value and boundary features are used as the basic inputs for calculating the subsequent disturbance intensity assessment value, which is used to support the construction of the stratigraphic change level in the candidate area map of sample collection.

[0008] In a preferred embodiment, the perturbation intensity assessment method is used when performing unit-level evaluation through multi-factor fusion, specifically including the following steps: Based on four types of variation factors, namely lithological variation factor, fault structure factor, stress response factor and historical migration factor, the original values ​​of the four types of variation factors are calculated for each spatial unit. The original values ​​of the four types of change factors are standardized so that they reflect the relative intensity of the corresponding geological change characteristics under a unified dimension. The standardized four types of change factors are converted into continuous hierarchical functions. The hierarchical functions are used to characterize the possibility of change and form the disturbance response expression corresponding to the spatial unit. Based on the disturbance response expression of all spatial units, a disturbance intensity assessment value is formed by using a non-equilibrium degree accumulation method, so that the assessment value can reflect the variation clustering, transition and structural continuity of different spatial units within the region. Based on the distribution of disturbance intensity assessment values ​​throughout the region, the change level is determined by a tiered threshold or statistical interval method, which distinguishes between low-change, medium-change, and high-change areas, and a candidate area map for sample collection is constructed according to preset standards.

[0009] In a preferred embodiment, the calculation steps for the disturbance intensity assessment value are as follows: Sensitivity weights are set according to the parameter sensitivity of the four types of change factors, and the four types of change factors are weighted and summed to obtain the initial value of the unit disturbance; The perturbation gradient direction in the spatial grid is used to perform neighboring cell difference value diffusion. Preset direction weights and distance attenuation weights are added during the diffusion to make the perturbation propagate continuously in space. The cumulative diffusion value is superimposed on the initial perturbation value of each cell to obtain the final perturbation intensity assessment value.

[0010] The preprocessing step of the candidate region map for sample collection before generating the sampling path group further includes: By jointly filtering terrain data, operational accessibility data, and equipment accessibility data within the region, a work space filter map is constructed, which excludes areas that do not meet the operational conditions during the path construction process.

[0011] In a preferred embodiment, during the construction of the sampling path group, the spatial connectivity between paths is evaluated. The spatial connectivity is formed by weighting the distance between the starting and ending points of the paths, the change value of the terrain slope difference, and the equipment accessibility score according to a preset influence coefficient, and the directional conversion and unification are performed before weighting. A weighted directed graph structure is established between any two paths in the path group. Each path is treated as a graph node, and the directed edges between nodes represent the path connection relationship. The edge weight is composed of the path connectivity score and the path switching complexity index. The path switching complexity index is modeled with reference to the shortest line switching time and is scored in segments based on equipment redeployment time, terrain adaptation time, and sample storage handover delay. The minimum weight Hamiltonian path generation strategy in graph search is adopted to find the path sequence that covers all path nodes and has the minimum sum of edge weights in the graph structure, so as to achieve the global optimal solution for path connectivity and execution reachability. In the generated optimal path sequence, identify the actual geographical areas corresponding to all path connection segments, obtain the risk exposure segments in the path, prioritize avoiding such segments in the path layout, and if the coverage is incomplete after avoidance, retain the path with the minimum risk strategy and generate corresponding mitigation suggestions, so that the path system can improve the overall operational safety while satisfying the spatial coverage integrity.

[0012] In a preferred embodiment, real-time comparison of sampled data with spatial reference data refers to: The sampling resistance variation, formation feedback waveform and sample physical properties were respectively constructed into continuous energy sequences, and then fused into a single comprehensive energy curve through standardization and geological feature sensitivity weighting. The single comprehensive energy curve is integrated stepwise to form an energy cumulative distribution curve, which is then compared layer by layer with the expected energy curve in the spatial reference data. The overall difference level is calculated by layer-by-layer difference superposition, and finally a formation response difference value is output to characterize the formation response degree of the sampling point.

[0013] In a preferred embodiment, adjusting the sampling path or sampling method based on the comparison results refers to: The stratigraphic response difference value of the current sampling point and the stratigraphic response difference values ​​of the two sampling points before and after it in the path sequence constitute a local neighborhood set. The neighborhood statistical distribution is formed by combining the mean and standard deviation. If the difference value of the current sampling point is greater than the mean of the neighborhood plus twice the standard deviation, or the difference between the current difference value and the adjacent previous or subsequent point exceeds the set mutation threshold, and the mutation shows a unilateral jump trend in the three point sequences, then it is considered that the sampling point does not conform to the continuity of the strata in the neighborhood. When the continuity of the sampling point with the adjacent stratigraphy is inconsistent, at least one of the following adjustment actions is triggered: redistribute the direction of the current path segment so that it extends along the trend direction of the adjacent point; switch the sampling mode of the current point to multi-layer penetration; insert intermediate sampling points between the current point and the adjacent points to supplement data integrity.

[0014] The technical effects and advantages of this invention are as follows: This invention acquires lithological characteristics, fault distribution, underground stress information, and historical sampling data of the target area, and fuses this multi-source geological information to form spatial reference data reflecting the degree of stratigraphic structural change, thereby achieving a multi-dimensional expression of regional geological structural differences. This fusion method not only overcomes the limitations of single geological parameters in reflecting true stratigraphic change characteristics but also effectively enhances the ability to identify geological anomaly areas. By dividing the entire region into multiple regional units with clearly defined geological boundaries based on spatial reference data, the identification of stratigraphic change levels has high spatial accuracy and geological interpretive significance, thus significantly improving the scientific rigor of geological information modeling in the pre-sampling stage and providing stable and reliable data support for downstream sampling path construction and sampling method matching.

[0015] In the process of generating sampling path groups, this invention dynamically optimizes the spatial layout of the paths based on the spatial block disturbance distribution marked in the sample collection candidate area map, combined with potential intersection conflicts between paths and the safety operation requirements within the work area. This ensures that the final path group can comprehensively cover high-disturbance areas, guarantee the representativeness of the sampling area, and avoid high-risk intersections and areas lacking sampling conditions, achieving deep adaptation between the work path and the geological target. By comprehensively considering path intersection constraints and operational safety factors, the sampling paths not only meet technical accuracy requirements but also possess stronger on-site deployment rationality, thereby significantly reducing potential scheduling conflicts, equipment interference, and geological operation safety hazards that may occur during actual sampling operations.

[0016] In the actual sampling process, this invention records the changes in sampling resistance, geological response waveforms, and preliminary physical properties of the sample at each sampling point in real time. These three key data types are used to construct the geological response characteristics of the sampling point. Based on comparative analysis with spatial reference data, the geological consistency of the current sampling point is verified in real time. When the comparison results indicate a deviation between the current sampling path or method and the target geological characteristics, the system can adjust the path or change the sampling method promptly based on the trend of stratigraphic response changes, thus enabling the sampling strategy to have dynamic adaptability during execution. This mechanism ensures that the sampling task can continuously respond to minor changes in the stratigraphic structure, resulting in higher reliability of the final sample data in terms of spatial coverage, lithological integrity, and engineering utilization value, significantly improving the quality of basic data for ore body modeling and resource assessment. Attached Figure Description

[0017] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is a schematic diagram of a mineral sample collection optimization method based on stratigraphic change identification, as described in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Reference Figure 1 The following examples were obtained: Example 1: An optimized method for mineral sample collection based on stratigraphic change identification, comprising the following steps: acquiring lithological characteristics, fault distribution, underground stress information, and historical sampling data of the target area, and fusing them to form spatial reference data reflecting the degree of stratigraphic structural change, used to divide regional units of different change levels; achieving high-dimensional perception of the underground structure of the target area through the collaborative fusion of multi-source geological data. Lithological characteristics reveal material composition and genetic environment, fault distribution reflects structural movement trends, underground stress information reveals stress state and fracture risk, and historical sampling data provides empirical correction reference. After uniformly converting the above information to the same scale and unit, spatial reference data is constructed, providing a unified evaluation basis for subsequent judgment of the degree of stratigraphic structural change. The generation of spatial reference data enables regional units to have unified comparability and distinguishability in subsequent stratigraphic disturbance analysis, serving as the basic data source for subsequent mapping and path planning.

[0020] Based on spatial reference data, a stratigraphic variation level distribution map is generated. By modeling the degree of geological disturbance within regional units, spatial blocks with different levels of disturbance are marked, constructing a candidate area map for sample collection. Building upon the spatial reference data generated in the previous step, a disturbance degree modeling process is established for regional units. By quantifying disturbance indicators, a continuous stratigraphic variation level is formed, thus generating a stratigraphic variation level distribution map. This map clearly identifies stable zones, transition zones, and strongly disturbed zones in the geological structure. These are then mapped onto spatial blocks to form a candidate area map for sample collection, providing an objective basis for prioritizing sampling areas. The generation of this map not only reflects the spatial characteristics of geological evolution but also provides a direct basis for optimizing sampling point settings, allowing sampling strategies to be more focused on key change areas.

[0021] Based on the candidate sample collection area map, sampling path groups are generated. During path generation, the spatial layout of the paths is adjusted according to the distribution of geological disturbance levels between spatial blocks, combined with the probability of path intersections and operational safety requirements. This ensures that the path groups meet operational safety conditions while maintaining effective coverage. The candidate sample collection area map generated in the previous step is utilized in a structured manner, and the spatial organization of the sampling plan is achieved by constructing sampling path groups. Disturbance level distribution is introduced as a spatial priority weight during path planning to guide paths to cover key areas. Path intersection analysis and operational safety boundary checks are used to dynamically adjust the path layout. This approach avoids path overlap and inefficient repeated sampling, while improving equipment operation safety and scheduling efficiency, thus achieving optimal coverage and safety.

[0022] Each sampling path in the sampling path group is matched with a corresponding sampling method to adapt the sampling method to different regional characteristics. A sampling method selection matrix is ​​established by coupling the spatial location attributes of the sampling path with three key parameters: geological disturbance level, burial depth, and topographic slope. Sampling methods include shallow sampling, borehole sampling, trenching, and pitting. This step dynamically matches the appropriate method, thereby improving sampling quality and field operation efficiency. This mechanism effectively avoids data distortion or sample failure caused by inappropriate sampling method selection, ensuring a reasonable match between sampling depth and method under different geological conditions, enhancing data reliability and the stability of the analytical basis.

[0023] During sample collection, the sampling resistance changes, geological response waveforms, and preliminary physical properties of the samples at each sampling point are recorded in real time and compared with spatial reference data. Based on the comparison results, the sampling path or sampling method is adjusted to ensure consistency between the sampling process and geological change characteristics. A dynamic closed loop is formed between the sampling process and the previously constructed spatial reference data. A multi-dimensional observation sequence is created through real-time acquisition of resistance changes, response waveforms, and physical properties, and a difference analysis is performed between this sequence and the expected stratigraphic structure characteristics. If a significant deviation is found between the actual sampling data and the stratigraphic reference, adjustments such as path fine-tuning, sampling method switching, or insertion of compensating sampling points are triggered. This ensures that the actual sampling process always conforms to real geological conditions, minimizing sampling errors and omissions, and achieving closed-loop control and intelligent optimization throughout the entire process.

[0024] To construct spatial reference data that accurately reflects the geological structure changes in the target area, it is first necessary to fuse and analyze multi-source geological information. This implementation proposes a spatial reference data generation method with high expression accuracy and geological structure fidelity. This step can also be performed using other processing methods in the prior art to achieve the same purpose, and is not limited to specific content, nor is it the core feature of this invention. The present invention provides a specific processing flow as follows: Multiple geological parameters are converted to the same scale, and amplitude normalization is used to make data from different sources comparable at the same order of magnitude. In this step, four types of original data sources are acquired: lithological characteristic data, fault distribution data, underground stress information, and historical sampling data. Lithological characteristics are represented by classified lithofacies codes, fault distribution is represented by fracture density values, underground stress information is numerically processed using local principal stress tensors, and historical sampling data includes sampling location, sample composition, and error residual values. Due to the different dimensions and significant differences in distribution span among the various data dimensions, amplitude normalization is adopted, that is, maximum and minimum value scaling is performed on each type of parameter to standardize them to the [0,1] interval, so that they have a uniform response scale. For example, the lithological coding sequence {3,7,5,1} is uniformly processed into {0.33,1.00,0.67,0.00}. In this way, data from different sources can be compared and fused within a unified index space, establishing a numerically compatible basis for cross-attribute parameters.

[0025] Geological multidimensional correlation analysis was used to determine the joint variation trend among parameters. Based on the normalized data, principal component analysis (PCA) and correlation matrix analysis were performed to measure the co-evolution characteristics among different types of geological parameters. In practical processing, the Pearson correlation coefficient can be used to assess the response relationship between lithological variation and stress gradient, identifying which parameters exhibit coupled fluctuation patterns in space. Taking data from a certain mining area as an example, the analysis revealed a high correlation of 0.81 between fault density and underground stress variation coefficient, indicating synchronous disturbance behavior between the two in local areas. Based on the correlation threshold, statistically significant parameter pairs were selected to construct a joint variation model, providing a coupling reference for subsequent spatial division.

[0026] Based on this, the target area is divided into multiple continuous spatial units using a multi-level partitioning approach, so that each spatial unit can serve as the basis for subsequent change level determination. During the spatial partitioning process, a geological feature continuity threshold can be introduced for hierarchical classification. A sliding window method is used to spatially scan the target area, constructing local parameter aggregation zones with a radius of one kilometer. The standard deviation and mean difference of parameters in each region are statistically analyzed, and similar regions are grouped into spatial units of the same level using a clustering algorithm. Each spatial unit is numbered using a fixed grid, and its internal data structure includes four types of parameters at a uniform scale: lithology, faults, stress, and historical deviation, providing a basic geographic structure for subsequent disturbance level modeling and distribution mapping.

[0027] In multidimensional fusion processing, the direction of geological change is obtained through trend line extraction, and the overall variability within the region is obtained through multi-point joint covariance estimation, enabling the spatial reference data to comprehensively reflect the structural differences in geological changes within the region. Trend line extraction uses each spatial unit as the center, analyzing the dominant direction of parameter changes along the X and Y axes, and fitting the spatial trend of geological parameters within the region using the least squares method to obtain multiple trend axes. The joint covariance estimation method uses each trend axis as the analysis path to model the spatial variability of the parameter set, obtaining the distribution curve of the variation amplitude between different spatial units. For example, a continuous increase in covariance value within a certain segment indicates a sudden change in geological parameters, providing reliable input for subsequent disturbance intensity modeling and stratigraphic change classification. The final spatial reference data not only maintains the original geological structure characteristics but also possesses a sensitive identification capability for minor geological disturbances, serving as a core reference for subsequent sample collection candidate area maps and path planning.

[0028] When generating a stratigraphic variation level distribution map, the steps for modeling the disturbance degree of each spatial unit include: collecting information on lithological variation intensity, fault structure characteristics, stress field response trends, and historical sampling offset within the spatial unit; constructing four types of variation factors; and calculating the disturbance modeling input data through multi-factor joint analysis. The four types of variation factors are expressed as standardized values ​​and used as disturbance modeling input vectors. To ensure that the input parameters in the disturbance degree modeling process have practical physical meaning and quantifiability, the original values ​​of the following four types of variation factors need to be accurately calculated. The four types of variation factors include lithological variation factor, fault structure factor, stress response factor, and historical offset factor corresponding to lithological variation intensity, fault structure characteristics, stress field response trends, and historical sampling offset information, respectively. Their specific calculation methods are as follows: The calculation of the original value of the lithological variation factor for lithological variation intensity is as follows: Each spatial unit is used as the sampling area for a geological profile. First, the number of lithological types within the unit is counted, and the lithological types appearing in the area, such as sandstone, shale, and limestone, are recorded. Then, they are classified according to their spatial continuity. If a certain lithology appears only locally and has a low area proportion, it is included as a high variation factor. Next, several uniform subdivision zones are divided along the main direction and perpendicular direction of the unit. The number of lithological category changes in each zone is calculated. The sum of the number of lithological changes in each subdivision zone is divided by the number of subdivisions to obtain the average frequency of change in the unit subdivision zone, which is used as the original value of the lithological variation intensity.

[0029] The calculation of the original values ​​of fault structure factors for fault structure characteristics is as follows: In each spatial unit, firstly, the fault density is calculated, which is the length density value obtained by dividing the total length of faults within the unit by the area of ​​the unit; secondly, the total number of fault intersections within the spatial unit is counted and recorded as the intersection number; thirdly, all fault strike angles are obtained, and the difference between the frequency of the most common strike angle and the frequency of the least common strike angle is calculated and defined as the strike variation amplitude value. To unify the scale characteristics of the three indicators, the fault density, intersection number, and strike variation amplitude value are linearly normalized to a maximum and minimum range, so that their values ​​are distributed between zero and one. Then, the three normalized indicators are combined according to the following steps: First, the average value of the fault density and the intersection number is taken to represent the structural complexity; then, this average value is added to the strike variation amplitude value and divided by two to obtain the final original values ​​of the fault structure characteristics.

[0030] Calculation of the original values ​​of the stress response factor for stress field response trends: From the geostress data obtained through borehole cores or stress testing instruments, the direction and amplitude variations of the principal stresses are extracted. Within each spatial unit, multiple sets of stress measuring points within a representative depth range are first selected, and the angles of change of the principal stress directions in the longitudinal and transverse directions are calculated. Then, the absolute differences in the principal stress directions between each measuring point are averaged to obtain the fluctuation level of the principal stress directions in that area. Furthermore, the variation trend of the principal stress amplitude is evaluated, the difference between the maximum and minimum principal stresses is calculated, and compared with the average principal stress level of the entire area to determine the strength of the stress field response in that unit.

[0031] The calculation of the original value of the historical offset factor for historical sampling offset information is as follows: In each spatial cell, firstly, the spatial offset distance between all historical sampling points and the current modeling prediction point within the cell's coverage area is calculated. The horizontal distance difference and vertical depth difference are calculated separately, and the average offset value of all samples is defined as the average offset distance of the cell. Secondly, the offset direction vectors of all historical sampling points are projected and classified into eight equal-angle sectors. The number of sampling points in each sector is counted, and the proportion of sampling points in the sector with the most points is calculated. This proportion is the directional consistency metric, reflecting whether the sampling offset direction within the cell shows a concentration trend. To form the original value of the historical sampling offset for this cell, the following method is used for combination: First, the average offset distance is normalized to between zero and one, representing the offset intensity; then, the directional consistency metric is also normalized to between zero and one, representing the offset concentration; finally, the two are multiplied to obtain the joint expression value.

[0032] In the perturbation modeling process, multi-factor fusion is used for unit-level evaluation, and gradient identification is performed at the boundaries of spatial units to extract the characteristics of abrupt geological structural boundary changes. Based on the initial values ​​of unit perturbations, a perturbation response field is constructed for all spatial units. For adjacent spatial units in this field, the difference between their initial values ​​of unit perturbations is calculated to generate a perturbation gradient field. When performing boundary identification, the spatial continuity of the gradient change rate is analyzed. A first-order partial derivative combined with a sliding window convolution method is used to identify the directionality of the perturbation gradient field in the spatial grid. The identification directions include the main northeast, northwest, east-west, and north-south axes to determine whether there are abrupt critical points. If the initial value of unit perturbation in a certain boundary direction shows a significant abrupt increase in three consecutive units (e.g., adjacent values ​​are 0.23→0.78→0.85 respectively), then that direction is identified as a possible boundary for abrupt geological structural changes.

[0033] The geological structural abrupt change boundary features obtained from gradient identification are integrated with the initial values ​​of unit perturbations to form a comprehensive perturbation expression containing structural jump information. These initial perturbation values ​​and boundary features are then used as the basis for calculating subsequent perturbation intensity assessment values, supporting the construction of stratigraphic variation levels in the sample collection candidate area map. During the construction of the sample collection candidate area map, all spatial units are classified into perturbation intensity levels based on the comprehensive perturbation expression model. Combined with boundary abrupt change information, a continuous but heterogeneous perturbation level zoning map is constructed. In this map, the central perturbation value and boundary abrupt change points jointly determine the range of variation levels in the region. For example, if multiple spatial units in a region exhibit high-intensity perturbation responses and the abrupt change gradients at the boundaries are consistent, it is identified as a "high-variation-level area," and this region is prioritized for inclusion in the sample collection priority area. If a region has low perturbation values ​​and no obvious structural abrupt changes at the boundaries, it is classified as a "low-variation-level area" and can be set as a sampling restriction area. Through the above processing, the stratigraphic variation level distribution map is ensured to express geological continuity while also possessing the ability to identify abrupt changes.

[0034] When performing unit-level evaluation through multi-factor fusion, the perturbation intensity evaluation method is adopted, which includes the following steps: Based on four types of variation factors, namely lithological variation factor, fault structure factor, stress response factor and historical migration factor, the original values ​​of the four types of variation factors are calculated for each spatial unit; each factor is based on a raster as the smallest unit to ensure that the data has spatial continuity and a basis for comparison, and to provide a unified input structure for subsequent standardization processing and perturbation modeling.

[0035] To eliminate dimensional differences among the original data of various change factors, the original values ​​of all factors need to be standardized. The standardization method employs a min-max normalization approach, uniformly transforming all values ​​to a standard range between zero and one, ensuring comparability across the same numerical dimension. After standardization, to further express the changing trends of different factors in response to geological disturbances, a continuous grading function is used to map the standardized values ​​to disturbance response levels. The grading function is set based on historical data statistical ranges; for example, values ​​below 0.2 correspond to extremely low disturbance probabilities, 0.2 to 0.4 correspond to low probabilities, and so on, with values ​​above 0.8 representing extremely high disturbance levels. Each spatial unit calculates its disturbance response expression using four grading functions, forming a four-dimensional disturbance response vector to characterize its changes across various geological dimensions.

[0036] The disturbance response expressions of all spatial units are used as input, and a non-equilibrium accumulation operation is performed to form the final disturbance intensity assessment value for that unit. The core idea of ​​non-equilibrium is to measure the clustering and abrupt change of spatial disturbances. The calculation steps for the disturbance intensity assessment value are as follows: For each spatial unit, the disturbance expression results of lithological variation factor, fault structure factor, stress response factor, and historical migration factor are calculated separately, and corresponding sensitivity weights are set according to the parameter sensitivity of the four types of variation factors. The setting of the above sensitivity weights is adjusted in stages based on the verification results of historical geological events, typical regional tectonic evolution paths, and expert scores. Specifically, the frequency and contribution of each factor in triggering stratigraphic abrupt change events in typical tectonic zones are first statistically analyzed, and normalization is performed in combination with the degree of change of modeling residuals. Factors with high anomaly dominance are assigned higher sensitivity weights. For example, in tectonic compression sections with frequent fault activity, the sensitivity weight of the fault structure factor can be set to more than twice that of other factors to reflect its dominant position in geological disturbances. Then, the four factors are multiplied by their corresponding sensitivity weight values, and a summation operation is performed to obtain the initial value of the cell perturbation, which serves as the basic measurement result of the degree of spatial perturbation and provides a quantitative starting point for the subsequent diffusion process.

[0037] Based on the distribution of initial disturbance values ​​in the unit cells, a set of diffusion paths guided by the disturbance gradient is constructed along the spatial grid structure. In practice, each spatial unit is centered, and its adjacent four- or eight-directional units are selected to form a local adjacency set. The difference in initial disturbance values ​​between the current unit and each adjacent unit is calculated, and this difference is used as the disturbance gradient value. To improve the responsiveness of spatial diffusion to the actual geological propagation direction, a direction determination mechanism is introduced. The preferred diffusion direction for each unit is determined based on structural parameters such as stratigraphic strike, structural axis, and fault extension trend. A preset directional weight is applied to this direction for numerical amplification, reflecting the fact that geological disturbances are more likely to propagate along the principal stress direction. Simultaneously, considering the inconsistency in spatial distance between different adjacent units and the central unit, a distance attenuation weight is introduced to reduce the impact according to the actual geographical distance, thus decreasing the impact of long-distance diffusion and reflecting the spatial decay attribute of disturbance transmission.

[0038] The constructed perturbation gradient values ​​are modified using both direction and distance factors to form perturbation diffusion values, which are then accumulated sequentially according to the regional spatial index. Specifically, starting from the target cell, perturbation diffusion values ​​are superimposed onto the initial perturbation values ​​of adjacent cells in the preferred diffusion direction and its secondary directions. This process is repeated with new adjacent cells as the center until the spatial gradient value decays below a preset threshold. During each propagation, the diffusion path is recorded and dynamically adjusted to ensure that the perturbation propagation conforms to the actual stratigraphic movement trend and geological boundary constraints, thereby simulating the spatial propagation chain of real stratigraphic perturbations. After all diffusion operations are completed, each cell will contain its original initial perturbation value and the sum of diffusion values ​​propagated from other cells, forming a complete perturbation response expression reflecting the local perturbation impact and the regional perturbation convergence effect.

[0039] The accumulated values ​​after diffusion are used as the final disturbance intensity assessment value for each spatial unit, and spatial classification is based on this value. In practical processing, by constructing a cumulative frequency distribution curve of regional disturbance intensity, the distribution critical points of different disturbance intervals are extracted to determine the boundaries between low, medium, and high disturbance. For example, the quantile method can be used to classify the top 30% of the disturbance intensity assessment values ​​as low disturbance areas, 30% to 70% as medium disturbance areas, and above 70% as high disturbance areas. Each level of spatial unit is assigned an independent code on the map, forming a candidate area map for sample collection used for subsequent route planning and sampling method selection. Through this disturbance intensity assessment method, a comprehensive expression of the aggregation, transition, and structural continuity of geological changes can be achieved, providing scientific and stable geological data support for mineral sample collection. After classification, the spatial units of each level are marked on the map, and a stratigraphic change level distribution map is generated. Based on this map, spatial blocks are formed by aggregating continuous units with the same disturbance level, and finally a sample collection candidate area map is constructed, providing a complete and dynamically variable target area guide for the subsequent layout of sampling paths.

[0040] In a preferred embodiment of the present invention, to ensure that the construction of the sampling path group can effectively avoid areas where operations cannot be carried out and to improve the actual accessibility and execution efficiency of path generation, the candidate area map for sample collection needs to be preprocessed before generating the sampling path group. Specifically, this includes the following steps: acquiring high-precision terrain data of the target area and performing slope analysis and geomorphic zoning extraction on the data. Statistical analysis of the slope value of each spatial unit is performed using a digital terrain model to identify units with slopes exceeding a set critical value, such as steep slopes greater than 30 degrees, landslide boundary areas, etc., as a preliminary set of excluded geomorphic units. Simultaneously, combining the elevation fluctuation range and hydrological confluence path identification results, valley erosion areas and potential landslide occurrence areas are marked, so that subsequent path construction avoids entering unstable, high-risk terrain sections.

[0041] Accessibility data is incorporated into the decision-making logic. This data includes existing road distribution, feasibility analysis of locations suitable for constructing temporary transport routes, personnel access radius restrictions, and hard-to-access areas marked in historical operation records. Using access radius analysis, an accessibility buffer zone is constructed for any given point, and blind spots are identified using a ground obstacle distribution map. Based on this, areas lacking ground accessibility or with excessively high access costs are marked as prohibited areas, and logical cascading rules are established to cover gray-area boundaries not completely excluded from slope data.

[0042] Accessibility data is used to verify the accessibility of each sampling area unit. This data includes performance parameters such as the sampling equipment's dimensions, maneuverability, traction requirements, and minimum turning radius. A three-dimensional operational corridor determination model is established using existing technology. The equipment's movement path is projected onto the terrain to identify spatial areas where the equipment cannot pass due to limitations in turning angle, width, or ground pressure. If a spatial unit exhibits a dead angle, excessive slope, or insufficient width in any movement path simulation, it is determined to be an inaccessible unit. This data result is further cross-validated and used for supplementary elimination based on the previous two steps.

[0043] The three types of data mentioned above are spatially filtered and then jointly overlaid to form a work space filter map. This filter map retains all spatial units that meet the requirements of terrain stability, accessibility, and equipment accessibility, while eliminating spatial units that do not meet any of these conditions. In the specific generation process, a spatial raster overlay judgment strategy is adopted, and Boolean operation rules in three-dimensional coordinates are set to assign a work feasibility identifier to each spatial unit. Only when all filtering conditions are met simultaneously is the unit marked as a valid area that can be used for sampling path construction. The final work space filter map serves as one of the basic inputs to the path generation algorithm, effectively excluding regional units that do not meet the work conditions, ensuring that the sampling path group has higher work feasibility, safety reliability, and actual construction matching during the planning process.

[0044] During the construction of the sampling path group, the spatial connectivity between paths is evaluated. Spatial connectivity is formed by weighting the distance between the path's starting and ending points, the change in terrain slope, and the equipment accessibility score according to a preset influence coefficient. Before weighting, directional transformation and standardization are performed. In practice, the spatial coordinates of the starting and ending points of each sampling path are extracted, and the endpoint connection distances between each path and adjacent paths are calculated as the distance between the starting and ending points. The change in terrain slope can be fitted using an elevation model of the path endpoint connection section to obtain the slope variation range, which is further quantified as the terrain adaptation cost required for spatial connectivity. The equipment accessibility score is formed by comprehensively considering the operational accessibility near the path endpoints, the ease of equipment transportation, and historical access records. The above three data points need to undergo directional transformation before weighting. When calculating spatial connectivity, the relevant indicators include: distance between the starting and ending points of a path (shorter is better, negative indicator); terrain slope change value (smaller value indicates better accessibility, negative indicator); and equipment accessibility score (higher is better, positive indicator). These indicators are semantically inconsistent in their positive and negative directions. Without first unifying their directions, direct weighting can lead to logical errors, such as a larger distance actually increasing the score. Only after all indicators have been converted to positive indicators where "larger values ​​indicate better path connectivity" can the weighting be reasonable. Otherwise, it will introduce interfering terms, causing the score to contradict the actual connectivity quality judgment. For example, negative indicators can be transformed in the following ways: by taking their reciprocal or inverse proportion, such as: distance conversion value = maximum distance / current distance; slope change conversion value = a constant / current slope change; or by using linear normalization to make them "larger values ​​indicate better". After unifying the direction, preset influence coefficients are set based on actual operational experience, and the three indicators are weighted and combined to form a quantitative result for evaluating the spatial connection efficiency between paths, which is used to guide subsequent path ranking.

[0045] A weighted directed graph structure is established between any two paths in the path group. Each path is treated as a graph node, and the directed edges between nodes represent path connections. The edge weights are composed of a path connectivity score and a path switching complexity index. The path switching complexity index is modeled using the shortest route change time model, and is scored in segments based on equipment redeployment time, terrain adaptation time, and sample storage handover delay. In the specific construction, paths are numbered as graph nodes, and a directed connection edge is established between every two path nodes. This connection edge represents the comprehensive cost required to seamlessly connect to the next path after the previous path is completed. The path connectivity score is the spatial connectivity quantification result obtained in the previous step. In the specific calculation of the path switching complexity index, it is modeled based on the actual operational behaviors required for equipment movement: the equipment redeployment time can be simulated by the equipment model and site constraints to measure the duration of loading, unloading, transfer, and re-erection; the terrain adaptation time depends on the slope, surface material, slipperiness, and other geological factors of the path connection section to assess the handling difficulty; and the sample storage handover delay can be given a precise range based on the sample sealing requirements, cold chain transfer connection time, or manual handover efficiency. The above assessments can be conducted using either rule-based scoring according to preset rules or expert-assigned scoring. The three scoring results are categorized and scored in stages, forming a path switching complexity index. Finally, the directed edge weights of the graph structure are constructed using the path connectivity score and the path switching complexity index, achieving an abstract model of the multi-path connection state.

[0046] This study employs a minimum-weight Hamiltonian path generation strategy from graph search to find the path sequence that covers all path nodes and has the minimum sum of edge weights within the graph structure, achieving a globally optimal solution for both path connectivity and execution reachability. In practical application, a pre-constructed weighted directed graph is input into a path combination model based on the Hamiltonian path search strategy. This model traverses all possible path permutations and combinations, comparing their sum of edge weights in the directed graph, while ensuring that each path node is visited only once. Breadth-first search or branch-and-bound algorithms are used for pruning optimization to reduce the combinatorial state space. Path validity verification constraints are introduced to eliminate path combinations that do not conform to spatial connectivity logic. Among all generated valid path sequences, the one with the minimum sum of edge weights is selected as the optimal solution for the sampling path execution order. This strategy can establish a minimum-cost closed-loop sequence among multiple path nodes, avoiding waste of operational resources and improving on-site operational efficiency and execution reachability. This optimal path sequence serves as a reference basis for subsequent layout optimization and safety assessment, ensuring that path sorting is both efficient and operationally feasible.

[0047] The system identifies the actual geographical areas corresponding to all path connectors in the generated optimal path sequence, identifies risk-exposed sections within the paths, and prioritizes avoiding such sections in the path layout. If incomplete coverage occurs after avoidance, the path is retained using a minimum-risk strategy, and corresponding mitigation recommendations are generated. This ensures that the path system improves overall operational safety while maintaining spatial coverage integrity. Specifically, the identification method involves projecting path connectors onto a regional geological hazard risk layer through spatial overlay analysis, including spatial boundaries such as landslide-sensitive areas, debris flow channels, and active fault zones, to determine whether the path traverses high-risk exposure sections. For risky path connectors, priority is given to adjusting path points to avoid the risk areas. If spatial gaps are created after avoidance, a minimum-risk strategy is activated, retaining the risky sections but requiring mitigation recommendations. This maximizes the safety and feasibility of the path layout without sacrificing spatial coverage integrity, providing a scientific basis for field sampling decisions.

[0048] In practical implementation, to achieve real-time dynamic adjustment of sampling paths or methods, it is necessary to comprehensively analyze the sampling resistance changes, formation feedback waveforms, and sample physical properties obtained during the sampling process, and continuously compare them with the expected geological response characteristics in spatial reference data. This comparison process requires not only consistency in the expression of the original data but also equivalence and comparability in their numerical structure. Therefore, it is implemented through the following steps: At each sampling point, the sampling resistance changes, formation feedback waveforms, and sample physical properties are collected sequentially to construct three sets of continuous data sequences, corresponding to the mechanical response, electromagnetic response, and physical property reflection during the sampling process, respectively. To ensure that these three types of data have consistent expressive power in numerical representation, they are first standardized to be normalized to the same dimension range; then, sensitivity weighting is applied based on the sensitivity of different parameters to changes in historical geological distribution. For example, in high-strength rock layers, the sampling resistance changes are more sensitive to formation differences, so the weight of this parameter is increased accordingly. Through the above processing, the three energy sequences are merged into a single comprehensive energy curve to express the overall geological response characteristics of the sampling point.

[0049] For this comprehensive energy curve, an energy cumulative distribution curve is generated through a stepwise integration method. This involves accumulating the energy change trends over sampling time or depth in integral form, so that the energy performance at each point is no longer limited to single-point data but forms a continuous energy trajectory. This trajectory can reflect the structural differences in energy input and absorption within the strata. To achieve comparison with the expected curve in the spatial reference data, equally spaced stratified nodes are set on the cumulative distribution curve, and the corresponding stratified energy values ​​are extracted.

[0050] The energy values ​​after stratification are compared layer by layer with the expected energy curve at that location in the spatial reference data. The overall difference level is calculated by summing the differences between each layer. For example, if a sampling point has a difference between the measured value and the expected value in each of the five stratification nodes, the differences are accumulated sequentially according to the stratification level to form the stratigraphic response difference value for that sampling point. This difference value is used to characterize the degree of agreement between the sampled data and the modeling expectations, and thus reflects whether there are potential geological abrupt changes or structural misjudgments.

[0051] The difference in the stratigraphic response at the current sampling point is combined with the difference between the two sampling points preceding and following it along the path to form a local neighborhood set. The neighborhood statistical distribution is constructed using the mean and standard deviation. If the current difference exceeds the neighborhood mean plus twice the standard deviation, or if the difference between the current point and its preceding or following point exceeds a set abrupt change threshold, and this abrupt change occurs only in a single direction, then the point is determined to be inconsistent with the stratigraphic continuity of its neighborhood. When a sampling point is inconsistent with the stratigraphic continuity of its neighborhood, a preset adjustment action is triggered. This preset adjustment action includes at least one of the following: first, repositioning the current path segment and extending it spatially towards the trend direction of neighboring points; second, switching the sampling method of the current point to multi-layer penetration sampling to detect potential changes below or around it; third, inserting intermediate sampling points between the current point and adjacent points to improve data resolution and interpretability. Through these steps, not only is adaptive optimization of the sampling path achieved, but also a high degree of consistency between the sampling data and the geological model is ensured, improving the reliability of the final mineral analysis and stratigraphic modeling.

[0052] The above algorithms or formulas are all dimensionless and numerical calculations, and the results are obtained by software simulation based on a large amount of collected data to obtain the most recent real-world results. The preset parameters are set by those skilled in the art according to the actual situation.

[0053] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0054] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0055] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0056] 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. An optimized method for mineral sample collection based on stratigraphic variation identification, characterized in that, Includes the following steps: The lithological characteristics, fault distribution, underground stress information and historical sampling data of the target area are acquired and fused to form spatial reference data that reflects the degree of change in stratigraphic structure, and is used to divide regional units of different change levels. Based on spatial reference data, a distribution map of stratigraphic variation levels is generated. By modeling the degree of geological disturbance within regional units, spatial blocks with different degrees of disturbance are marked, and a candidate area map for sample collection is constructed. Based on the candidate area map for sample collection, a sampling path group is generated. During the path generation process, the spatial layout of the path is adjusted based on the distribution of geological disturbance between spatial blocks, combined with the possibility of path intersection and operational safety requirements, so that the path group can meet operational safety conditions while ensuring coverage effectiveness. For each path in the sampling path group, a corresponding sample collection method is matched. During the sample collection process, the sampling resistance change value, geological response waveform and preliminary physical properties of the sample at each sampling point are recorded in real time and compared with spatial reference data. The sampling path or sampling method is adjusted according to the comparison results.

2. The method for optimizing mineral sample collection based on stratigraphic change identification according to claim 1, characterized in that, The steps for jointly processing lithological characteristic data, fault distribution data, underground stress information, and historical sampling data when generating spatial reference data further include: Multiple geological parameters are converted to the same scale, and the amplitude normalization method makes the data from different sources comparable at the same order of magnitude. The joint change trend between parameters is judged by geological multidimensional correlation analysis. On this basis, the target area is divided into multiple continuous spatial units in a multi-level partitioning method, so that each spatial unit can serve as the basis for subsequent change level determination. When performing multidimensional fusion processing, the direction of geological change is obtained by trend line extraction, and the overall variability within the region is obtained by multi-point joint covariance estimation.

3. The method for optimizing mineral sample collection based on stratigraphic change identification according to claim 1, characterized in that, When generating a geological change level distribution map, the steps for modeling the degree of disturbance in each spatial unit include: Information on lithological variation intensity, fault structure characteristics, stress field response trends and historical sampling offsets within spatial units is collected to construct four types of variation factors. The input data for disturbance modeling is calculated through multi-factor joint analysis. During the perturbation modeling process, unit-level evaluation is performed through multi-factor fusion, and gradient identification is performed at the spatial unit boundary to extract the boundary features of geological structural abrupt changes.

4. The optimized method for mineral sample collection based on stratigraphic change identification according to claim 3, characterized in that, When performing unit-level evaluation through multi-factor fusion, the perturbation intensity evaluation method is adopted, which includes the following steps: Based on four types of variation factors, namely lithological variation factor, fault structure factor, stress response factor and historical migration factor, the original values ​​of the four types of variation factors are calculated for each spatial unit. The original values ​​of the four types of change factors are standardized, and the standardized four types of change factors are converted into continuous hierarchical functions, and the disturbance response expression corresponding to the spatial unit is formed accordingly. Based on the disturbance response expression of all spatial units, a disturbance intensity assessment value is formed using a non-equilibrium degree accumulation method. According to the distribution of the disturbance intensity assessment value in the entire region, the change level is determined by a graded threshold or statistical interval method, so that the change level distinguishes between low change area, medium change area and high change area, and a sample collection candidate area map is constructed according to a preset standard.

5. The method for optimizing mineral sample collection based on stratigraphic change identification according to claim 4, characterized in that, The calculation steps for the disturbance intensity assessment value are as follows: Sensitivity weights are set according to the parameter sensitivity of the four types of change factors, and the four types of change factors are weighted and summed to obtain the initial value of the unit disturbance; The perturbation gradient direction in the spatial grid is used to perform neighboring cell difference value diffusion. Preset direction weights and distance attenuation weights are added during the diffusion to make the perturbation propagate continuously in space. The cumulative diffusion value is superimposed on the initial perturbation value of each cell to obtain the final perturbation intensity assessment value.

6. The optimized method for mineral sample collection based on stratigraphic change identification according to claim 5, characterized in that, The preprocessing step of the candidate region map for sample collection before generating the sampling path group further includes: A joint screening of terrain data, operational accessibility data, and equipment accessibility data within the region is used to construct an operational space filtering map.

7. The method for optimizing mineral sample collection based on stratigraphic variation identification according to claim 6, characterized in that, During the construction of the sampling path group, the spatial connectivity between paths is evaluated. The spatial connectivity is formed by weighting the distance between the starting and ending points of the path, the change value of the terrain slope difference, and the equipment accessibility score according to the preset influence coefficient. Before weighting, directional transformation and unification processing are performed. A weighted directed graph structure is established between any two paths in the path group. Each path is treated as a graph node, and the directed edges between nodes represent the path connection relationship. The edge weight is composed of the path connectivity score and the path switching complexity index. The path switching complexity index is modeled with reference to the shortest line switching time and is scored in segments based on equipment redeployment time, terrain adaptation time, and sample storage handover delay. The minimum weight Hamiltonian path generation strategy in graph search is adopted to find the path sequence that covers all path nodes and has the minimum sum of edge weights in the graph structure, so as to achieve the global optimal solution for path connectivity and execution reachability. Identify the actual geographical areas corresponding to all path connection segments in the generated optimal path sequence, obtain the risk exposure segments in the path, prioritize avoiding such segments in the path layout, and if the coverage is incomplete after avoidance, retain the path with the minimum risk strategy and generate corresponding mitigation suggestions.

8. The method for optimizing mineral sample collection based on stratigraphic change identification according to claim 6, characterized in that, Real-time comparison of sampled data with spatial reference data refers to: The sampling resistance variation, formation feedback waveform and sample physical properties were respectively constructed into continuous energy sequences, and then fused into a single comprehensive energy curve through standardization and geological feature sensitivity weighting. The single comprehensive energy curve is integrated stepwise to form an energy cumulative distribution curve, which is then compared layer by layer with the expected energy curve in the spatial reference data. The overall difference level is calculated by layer-by-layer difference superposition, and finally a formation response difference value is output to characterize the formation response degree of the sampling point.

9. The optimized method for mineral sample collection based on stratigraphic change identification according to claim 8, characterized in that, Adjusting the sampling path or sampling method based on the comparison results refers to: The stratigraphic response difference value of the current sampling point and the stratigraphic response difference values ​​of the two sampling points before and after it in the path sequence constitute a local neighborhood set. The neighborhood statistical distribution is formed by combining the mean and standard deviation. If the difference value of the current sampling point is greater than the mean of the neighborhood plus twice the standard deviation, or the difference between the current difference value and the adjacent previous or subsequent point exceeds the set mutation threshold, and the mutation shows a unilateral jump trend in the three point sequences, it is determined that the sampling point is inconsistent with the continuity of the neighboring strata. When the sampling point is inconsistent with the continuity of the neighboring strata, a preset adjustment action is triggered.

Images