Geological mapping method based on three-dimensional modeling of unmanned aerial vehicle

Abstract

The application discloses a geological mapping method based on three-dimensional modeling of an unmanned aerial vehicle, relates to the technical field of geological survey and intelligent path planning, and comprises the following steps: S1, an initial feature data set is extracted from geological information by constructing a three-dimensional space data processing framework, original data is decomposed into a plurality of sub-region data sets, and a preliminary classified geological information set is obtained; S2, according to the preliminary classified geological information set, abnormal values and boundary information in the sub-region data set are analyzed, and a potential key geological feature point set is determined; and S3, the point set is matched with geological environment parameters, and if the matching degree is lower than a preset threshold, the point is recalibrated, and an optimized geological feature distribution map is obtained. The geological mapping method based on three-dimensional modeling of the unmanned aerial vehicle effectively improves the accuracy and efficiency of geological survey in a complex terrain environment, and provides comprehensive technical support for field geological survey.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological survey and intelligent path planning, and particularly relates to a geological mapping method based on three-dimensional modeling of unmanned aerial vehicles. BACKGROUND

[0002] As an important basis for earth science research and resource exploration, geological survey is directly related to the strategic needs of national resource development and environmental protection, and its precision and efficiency have a key impact on scientific decision-making. In field surveys, reasonable planning of sampling points and paths is a core link to improve work quality.

[0003] However, the current method often fails to balance comprehensiveness and pertinence when faced with complex terrain and variable geological environment, especially in data integration and dynamic adjustment. Existing solutions rely heavily on manual experience, lack depth in mining complex three-dimensional spatial information, and lack real-time optimization capabilities, resulting in uneven resource allocation and limited survey efficiency when faced with large-scale, multi-target survey tasks. In this context, the field of geological survey faces significant technical challenges. The first problem is how to accurately identify key geological boundaries and abnormal areas from complex three-dimensional spatial data. These areas are often hidden in vast amounts of information and difficult to quickly locate through traditional means. Solving this problem further involves how to intelligently plan sampling point layout and navigation paths that can cover key areas and save time based on the identification results. The two factors are closely related, with the former determining the scientificity of the plan and the latter directly affecting the operability of field work. If these two problems cannot be effectively solved, survey work will be difficult to escape from the dilemma of low efficiency and high cost. SUMMARY

[0004] The purpose of the present application is to provide a geological mapping method based on three-dimensional modeling of unmanned aerial vehicles, which combines three-dimensional modeling results with geological survey needs, intelligently identifies key geological features, and optimizes sampling point layout and navigation paths.

[0005] To achieve the above purpose, the present application provides the following technical solution: a geological mapping method based on three-dimensional modeling of unmanned aerial vehicles, comprising:

[0006] S1, by constructing a three-dimensional spatial data processing framework, extracting an initial feature data set from geological information, decomposing the original data into a plurality of sub-regional data sets, and obtaining a preliminary classified geological information set;

[0007] S2, according to the preliminary classified geological information set, analyzing the abnormal values and boundary information in the sub-regional data set, and determining a set of potential key geological feature points;

[0008] S3, for the key geological feature point set, the point set is matched with the geological environment parameter, if the matching degree is lower than the preset threshold, the point is recalibrated, and the optimized geological feature distribution map is obtained;

[0009] S4, according to the optimized geological feature distribution map, the feature distribution map is grid analyzed, and the initial sampling point position set is determined;

[0010] S5, for the initial sampling point position set, the terrain environment under the passing cost is calculated, if the path segment cost is higher than the preset threshold, the alternative path is recalculated, and the preliminary optimized navigation path scheme is obtained;

[0011] S6, according to the preliminary optimized navigation path scheme, the real-time optimization mechanism is integrated, the path is adjusted, if the terrain obstacle information is detected, the path re-planning process is triggered, and the final navigation path is determined;

[0012] S7, for the final navigation path and the sampling point position set, the multi-target investigation task is divided into multiple subtask units, the task execution priority is obtained, and the field investigation execution scheme is obtained;

[0013] S8, according to the field investigation execution scheme, the potential risk point of the execution scheme is simulated and analyzed, if the risk value exceeds the preset threshold, the scheme is adjusted locally, and the final execution investigation plan is determined.

[0014] Preferably, the S1 comprises extracting the initial feature data set from the geological information by constructing a three-dimensional space data processing framework, adopting a multi-level grid division method to decompose the original data according to the data distribution characteristics in complex terrain environment, obtaining the sub-region data set and the classification information set; the original data is decomposed into multiple sub-region data sets by using the multi-level grid division method, the classification results of local geological information are obtained according to the data distribution characteristics in the sub-region set; according to the classification results of the sub-region data set, the classification information is optimized and adjusted by using support vector machine algorithm according to the data distribution characteristics in complex terrain environment, and the geological information set is determined; if there is uneven data distribution in the geological information set, the sub-region set is supplemented by data interpolation method, and the balanced distribution of geological information data set is obtained; according to the balanced distribution of geological information data set, the classification information is processed by using space interpolation technology according to the multi-level network structure in three-dimensional space, and the geological information distribution map is obtained; through the geological information distribution map, combined with the initial feature data set, it is judged whether there is abnormal data point, if there is abnormal data point, it is smoothed, and the modified geological information distribution result is obtained; according to the modified geological information distribution result, the classification information in the multi-level network is integrated by using data fusion technology according to the data characteristics in the terrain environment, and the final geological information comprehensive data set is determined.

[0015] Preferably, the S2 includes that the S2 includes the geological information and the preliminary classification results, the sub-region set is hierarchically analyzed by using a data processing tool, the distribution characteristics of the abnormal value points and the boundary information are obtained, and the preliminary abnormal distribution range is determined; according to the abnormal distribution range, the local feature comparison is performed on the abnormal value points and the boundary information in combination with a depth analysis method, and an abnormal region distribution map is obtained; through the distribution map, the preset threshold is used for screening processing on the abnormal region, and a candidate point set of key features is obtained; according to the candidate point set, the support vector machine algorithm is used for classification optimization on the point set in combination with the distribution law of the potential features, and the key geological feature point is determined; if the key geological feature point is unevenly distributed in the sub-region set, the point set is adjusted by using a data smoothing tool, and an evenly distributed feature point set is obtained; according to the evenly distributed feature point set, the point set is associated and integrated by using a space mapping technology according to the spatial characteristics of the geological information, and a comprehensive geological feature distribution map is obtained; through the comprehensive geological feature distribution map, the potential features are secondarily verified in combination with the intermediate results in the screening process, and a final key geological feature point set is determined.

[0016] Preferably, the S3 includes that a feature association framework is constructed by using a data integration method according to the corresponding relationship between the geological feature point set and the environmental parameters, and preliminary corresponding degree distribution data are obtained; the preliminary corresponding degree distribution data are screened by using a preset threshold, if the corresponding degree is lower than the preset threshold, the point set is corrected to obtain a corrected point distribution data set; according to the corrected point distribution data set, a matching deviation between the point set and the environmental parameters is analyzed by using a data comparison tool, and a point subset is determined; according to the point subset, the position of the point subset is calibrated by using a space adjustment method, and an adjusted point distribution structure is obtained; according to the adjusted point distribution structure, the corresponding degree between the point set and the environmental parameters is recalculated by using the feature association framework, and updated matching distribution data are obtained; according to the updated matching distribution data, if the matching distribution data still has points that do not reach the preset threshold, these points are secondarily optimized by using a data smoothing tool, and a final optimized distribution map is obtained; according to the final optimized distribution map, the point set and the geological features are integrated by using a space mapping technology, and a comprehensive geological feature distribution view is determined.

[0017] Preferably, S4 comprises, for the initial set of sampling point positions, using a spatial analysis tool to compare the sampling point positions with the distribution map of geological features, to determine the coverage proportion of the sampling point positions in the high-density area, and if the coverage proportion does not reach the preset threshold, adjusting the positions of the sampling points to obtain an adjusted sampling point distribution dataset; according to the adjusted sampling point distribution dataset, calculating the relevance of the sampling point positions and the feature distribution, analyzing whether the point distribution in the high-density area is balanced, and if the distribution deviation exceeds the preset range, locally optimizing the point positions to obtain an optimized point distribution structure; for the optimized point distribution structure, using a spatial interpolation method to process the feature distribution within the grid division to obtain regional analysis data and determine the boundary range of the density area; according to the boundary range of the density area, using a data screening tool to perform secondary verification on the sampling point positions to determine whether the point positions are all within the boundary range, and if there are point positions deviating from the boundary, adjusting the positions of the deviating point positions to obtain a corrected set of sampling points; for the corrected set of sampling points, using a data integration framework to associate and map the position data with the distribution characteristics of geological features to obtain a comprehensive sampling point distribution view; according to the comprehensive sampling point distribution view, using a visualization processing tool to superimpose and compare the sampling layout and the regional analysis results to determine a final sampling point position configuration scheme; for the final sampling point position configuration scheme, using a data storage module to archive the position data and the feature distribution information to obtain a structured sampling layout dataset.

[0018] Preferably, S5 comprises, for the preliminary optimized navigation path scheme, using a data processing tool to analyze the matching degree of the path scheme and the terrain, and if the matching degree is lower than the preset threshold, locally correcting the path scheme to obtain a corrected path distribution; according to the corrected path distribution, calculating the relevance of the sampling point positions and the navigation path, and if the relevance cannot cover the target area, redistributing the point positions to determine an adjusted point layout; for the adjusted point layout, using an information integration framework to fuse and process the path scheme and the passing cost data to obtain a comprehensive path cost distribution view; according to the comprehensive path cost distribution view, using a data screening tool to identify the cost segments in the navigation path that are higher than the preset threshold, and if the proportion of the cost segments that are higher than the preset threshold exceeds the preset proportion range, planning a replacement path for the cost segments that are higher than the preset threshold to obtain an optimized path structure; for the optimized path structure, using a spatial interpolation method to analyze the coverage range of the path scheme and the sampling point positions to obtain regional distribution data; according to the regional distribution data, using a data verification module to evaluate the compliance of the path scheme with the planning requirements, and if the compliance does not reach the preset standard, fine-tuning the path structure to determine a final navigation path configuration.

[0019] Preferably, the S6 comprises acquiring dynamic environment information from the field investigation scene, continuously monitoring changes in the environment, and obtaining the latest terrain and obstacle data distribution; according to the terrain and obstacle data distribution, using the pre-established path updating rule, comparing and analyzing the navigation path information, if the terrain obstacles affect the path passage is detected, triggering the path adjustment process, determining the preliminary adjusted path scheme; for the preliminary adjusted path scheme, through the spatial analysis tool, the matching degree of path information and dynamic environment data is calculated, if the matching degree is lower than the preset threshold, the path local section is modified, and the modified path layout is obtained; according to the modified path layout, using the information fusion framework, the environment change data collected in real time and the path information are integrated and processed, and the comprehensive environment adaptability distribution view is obtained; for the comprehensive environment adaptability distribution view, the paragraphs affected by the terrain obstacles in the navigation path are identified, if the proportion of the identified affected paragraphs exceeds the preset range, the re-planning process is started, and the alternative path scheme is determined; according to the alternative path scheme, using the spatial interpolation method, the path layout and the field investigation scene coverage range are calculated, and the path distribution data is obtained; for the path distribution data, the fit degree of the path scheme and the investigation scene demand is evaluated, if the fit degree does not reach the preset standard, the path layout is adjusted, and the final navigation path configuration is determined.

[0020] Preferably, the S7 comprises using the pre-established resource configuration model to analyze the spatial distribution relationship between the sampling position and the path according to the data of the navigation path and the sampling position, and obtaining the resource configuration preliminary allocation scheme; according to the preliminary allocation scheme, the distribution uniformity of the resource configuration is evaluated through the spatial analysis tool by fusing the balance principle, if the distribution uniformity is lower than the preset threshold, the resource configuration is locally adjusted, and the adjusted configuration layout is determined; for the adjusted configuration layout, a task allocation framework is constructed, a multi-target task is decomposed into a plurality of sub-task units, and a set of decomposed task units is obtained; according to the set of decomposed task units, the matching degree of the sub-task unit and the resource configuration layout is calculated by using the priority sorting method, and the task execution priority order is obtained; for the task execution priority order, the priority order and the overall demand of the investigation scheme are compared, if the comparison result shows that the task unit execution order does not conform to the overall demand, the priority order is adjusted, and the optimized execution sequence is determined; according to the optimized execution sequence, using the data verification tool, the fit degree of the execution sequence and the complete execution scheme is analyzed, and the final field investigation execution plan is obtained; for the final field investigation execution plan, through the data mapping method, the execution plan is associated with the actual distribution of the navigation path and the sampling position, and the execution path distribution is obtained.

[0021] Preferably, the S8 comprises a field investigation execution scheme, a preliminary risk point is identified by pre-establishing a risk assessment framework, obtaining regional distribution information, and determining a preliminary risk distribution map; according to the preliminary risk distribution map, the regional distribution information is deeply analyzed, if the analysis result shows that the risk value of the region exceeds the preset threshold, the specific risk points are extracted through the data screening tool to obtain a risk point set; for the risk point set, the influence range of each risk point is dynamically simulated through a simulation analysis tool, the boundary data of the influence range is obtained from the simulation result, and a risk influence range list is determined; according to the risk influence range list, a local adjustment mechanism is used to re-plan the part involving the risk point in the execution scheme, the related path and task allocation are adjusted through a spatial distribution optimization tool, and an adjusted scheme segment is obtained.

[0022] Preferably, the S8 further comprises, for the adjusted scheme segment, seamlessly connecting the adjusted scheme segment with other parts of the original execution scheme, if a task allocation conflict is found in the connection process, the conflict is resolved through a priority sorting tool, and an integrated complete scheme is obtained; according to the integrated complete scheme, a verification tool is used to comprehensively detect the executability of the scheme, if the detection result shows that there is an execution obstacle in the link, a substitute path is extracted through a backup scheme database, and a final investigation plan is determined; for the final investigation plan, the plan content is associated and matched with the actual environmental information of the field investigation, environmental adaptability feedback is obtained, and an optimized execution guide is obtained.

[0023] From the above technical solutions, the present application has the following beneficial effects:

[0024] The geological mapping method based on the three-dimensional modeling of the unmanned aerial vehicle extracts an initial feature data set from massive geological information by constructing a three-dimensional space data processing framework, decomposes the original data into a plurality of sub-region data sets by using a multi-level grid division method, identifies key geological features by using a feature extraction algorithm, constructs a feature correlation model to optimize the geological feature distribution map. According to the distribution map, a sampling point layout scheme is designed, a path optimization model is constructed in combination with the navigation path planning requirement, and a real-time optimization mechanism is integrated to cope with the dynamic changes in the field investigation. The present application further constructs a task allocation model in combination with the resource allocation balance principle, decomposes the multi-target investigation task into sub-task units, and embeds a scientific decision support module for risk analysis, and finally determines an executable investigation plan. The system can effectively improve the accuracy and efficiency of geological investigation in complex terrain environment, and provides comprehensive technical support for field geological investigation. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 The method flowchart of the present application. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] like Figure 1 As shown, the present invention provides a technical solution: a geological mapping method based on UAV three-dimensional modeling, comprising:

[0028] S1. By constructing a three-dimensional spatial data processing framework, an initial feature dataset is extracted from geological information, and the original data is decomposed into multiple sub-regional datasets to obtain a preliminary classified geological information set;

[0029] S2. Based on the preliminary classified geological information set, analyze the outliers and boundary information in the sub-regional data set to determine the potential key geological feature point set;

[0030] S3. For a set of key geological feature points, match the set of points with geological environment parameters. If the matching degree is lower than a preset threshold, recalibrate the points to obtain an optimized geological feature distribution map.

[0031] S4. Based on the optimized geological feature distribution map, perform grid analysis on the feature distribution map to determine the initial sampling point location set;

[0032] S5. For the initial set of sampling point locations, the travel cost under the terrain environment is taken into account. If the cost of the path segment is higher than a preset threshold, an alternative path is recalculated to obtain a preliminary optimized navigation path solution.

[0033] S6. Based on the preliminary optimized navigation path plan, the real-time optimization mechanism is integrated to adjust the path. If terrain obstacles are detected, the path replanning process is triggered to determine the final navigation path.

[0034] S7. Based on the final navigation path and sampling point location set, the multi-target survey task is decomposed into multiple subtask units, the task execution priority is obtained, and a field survey execution plan is obtained;

[0035] S8. Based on the field investigation implementation plan, conduct simulation analysis on the potential risk points of the implementation plan. If the risk value exceeds the preset threshold, make partial adjustments to the plan and determine the final implementation plan of the investigation.

[0036] The method is based on the constructed three-dimensional space data processing framework, uses the unmanned aerial vehicle to collect high-precision three-dimensional model data of the geological scene, and performs regional processing and preliminary classification on the original geological information. Through comprehensive analysis of abnormal geological data and boundary geological features in each sub-region, potential key geological points are automatically identified, and further compared with the geological environment parameters in the database to determine their rationality and reliability. If the matching degree is insufficient, the spatial calibration of the key points is performed using historical data and a geological evolution model, so as to obtain a more accurate geological feature distribution map. After obtaining the optimized distribution map, the survey area range is determined by setting a grid strategy, and the initial sampling point position is determined accordingly. Combined with the unmanned aerial vehicle path planning module, the passage cost in the sampling path is evaluated. If the cost of a path segment is higher than the set warning value, the system will automatically calculate an alternative path to improve the overall passage efficiency and safety. At the same time, the real-time optimization mechanism can dynamically correct the path during field sampling. Once a terrain obstacle is identified, the path re-planning process will be automatically entered, and the navigation path will be updated. After the navigation path is determined, the system divides the overall survey task into several sub-task units according to the final path and sampling point distribution, and generates task execution priorities according to resource limitations, time windows and other factors. Finally, the field survey execution plan containing task content, sampling order, path planning and resource allocation information is formed, and further combined with historical cases and a simulation engine to simulate and evaluate possible risks. If the risk exceeds the threshold, the system will dynamically adjust the local strategy to ensure the implementability and safety of the survey plan.

[0037] The present embodiment realizes efficient classification and analysis of geological information through three-dimensional modeling data collected by an unmanned aerial vehicle, improves the accuracy of geological feature recognition, effectively eliminates misidentified points through point matching and calibration mechanism with geological parameters, enhances the reliability of the geological distribution map, significantly improves the feasibility and safety of the navigation path through path cost evaluation and real-time adjustment mechanism, and reduces the sampling cost. The multi-task decomposition and priority sorting strategy improves the task execution efficiency, and the risk simulation and local adjustment strategy effectively ensures the controllability and safety of the implementation of the scheme.

[0038] S1 includes extracting an initial feature data set from geological information by constructing a three-dimensional space data processing framework, decomposing original data using a multi-level grid division method for data distribution characteristics in a complex terrain environment to obtain a sub-region data set and a classification information set; using a multi-level grid division method, the original data is decomposed into a plurality of sub-region data sets, and the classification results of the local geological information are obtained according to the data distribution characteristics in the sub-region set; according to the classification results of the sub-region data set, combining the data distribution characteristics in the complex terrain environment, the classification information is optimized and adjusted by using a support vector machine algorithm to determine the geological information set; if there is uneven data distribution in the geological information set, the sub-region set is supplemented by a data interpolation method to obtain an evenly distributed geological information data set; according to the evenly distributed geological information data set, for the multi-level grid structure in the three-dimensional space, the classification information is processed by using a spatial interpolation technique to obtain a geological information distribution map; through the geological information distribution map, combining the initial feature data set, it is judged whether there is an abnormal data point, if there is an abnormal data point, it is smoothed to obtain a modified geological information distribution result; according to the modified geological information distribution result, for the data characteristics in the terrain environment, the classification information in the multi-level grid is integrated by using a data fusion technology to determine the final geological information comprehensive data set.

[0039] In the embodiment, first, the target area is scanned by the unmanned aerial vehicle to obtain original geological data containing elevation, slope, lithology, topographic clues and other information. All data are unified according to a preset coordinate system and imported into a three-dimensional space data processing system. The first step of the system is to process the terrain data in stages to construct a multi-level grid structure. Specifically, the system divides the entire region into a plurality of primary main grids according to the undulating degree of the terrain, and the division standard of the main grid is that each grid covers a range of 100 meters by 100 meters in the horizontal direction, and the elevation change is not more than 50 meters. If the elevation change in a main grid exceeds 50 meters, the main grid is automatically subdivided into 10 by 10 sub-grids, each with a size of 10 meters by 10 meters, for more detailed geological information collection and analysis.

[0040] The data samples in each sub-grid are extracted and subjected to attribute analysis, with main parameters including lithology code, mineral type, structure direction, and fracture density. For these attribute data, the system uses a statistical module to calculate basic statistics such as mean, range, and standard deviation. On this basis, adjacent data points in each sub-grid are paired, the attribute difference between each pair of data is calculated, and the average of all differences is taken. If the average difference is lower than a preset similarity threshold, which is set to 10% of the total attribute variation range, the sub-grid is classified into the same category. The preliminary classification result is used as the initial input for support vector machine training.

[0041] The training samples of the support vector machine come from historical geological survey data, and the data set contains actual samples of multiple known lithologies, each of which is not less than 100. The system uses five-fold cross-validation technology to evaluate the performance of the trained model, and the target classification accuracy is set to not less than 95%. If the classification accuracy meets the standard, the model will be used to optimize the classification results of the current sub-grid. The optimization process includes fine-tuning of the classification boundary to maximize the maximum interval of the discriminant function between adjacent categories, thereby improving the consistency of classification.

[0042] The training process of the support vector machine includes six specific steps of data preparation, feature extraction, model initialization, training execution, model verification and parameter tuning, each of which has clear input, operation logic and determination condition, as follows: First, data preparation. The system extracts not less than 3000 sample data with complete labels from historical geological surveys, which contain information such as lithology type, stratigraphic structure, spatial position, etc., and cover not less than 5 main lithology categories in the area. Each sample needs to have a clear classification label, i.e. the lithology category it belongs to. The samples are evenly divided by category to ensure that each category has not less than 500 instances. Second, feature extraction. The system extracts a fixed number of input feature variables from each sample, which is set to 12 feature dimensions in this embodiment, including: relative position value of three-dimensional space coordinates, stratigraphic thickness, rock density, mineral composition ratio, porosity, color coding, weathering degree score, structure line inclination, fracture development index, average lithology code of adjacent samples, regional slope, and distance to the nearest fault zone, etc. All feature values are normalized by using maximum and minimum linear transformation to map them to 0 to 1 to eliminate the dimension effect. Third, initialize the support vector machine model. The model selects the radial basis function as the kernel function to enhance the classification ability of the non-linear boundary. The penalty parameter is set to 100 and the kernel function width parameter is set to 0.1, both of which are empirical initial values. Fourth, execute the training process. The whole sample data is divided into 80% as the training set and 20% as the validation set for model fitting. The training target is to maximize the interval and minimize the classification error. Gradient descent method is used to optimize the Lagrange function during training, with an upper limit of 10000 iterations and an early termination condition of a change in the objective function less than 0.0001. After training, the model parameters include the support vector set, the Lagrange multiplier, the bias term and the decision function expression. Fifth, model verification. The remaining 20% of the validation set is classified and predicted, the predicted label is compared with the true label, and the classification accuracy, precision, recall and F1 score are calculated. In this embodiment, the classification accuracy is set to not less than 95%, the precision and recall are set to not less than 93%, and the F1 score is set to not less than 94% as the model passing standard. Sixth, parameter tuning. If the model does not meet the above indicators, the system will adjust the penalty parameter and kernel width parameter in turn, use grid search method to cross-validate in the parameter interval, search the penalty parameter between 10 to 1000 and the kernel function width between 0.01 to 1, with search steps of 10 and 0.05 respectively. The parameter combination with the best overall performance is selected in the cross-validation, the model is retrained and verified until all evaluation indicators are met. After completing all the above steps, the system will solidify and save the trained model as a discrimination model for subsequent geological classification optimization steps, and perform high-precision prediction and automatic classification processing on the data of the sub-regions to be classified.

[0043] After the initial classification optimization, the system further detects the spatial density of data points within the subgrids. Specifically, the number of sampling points in each subgrid is calculated, and if the number of data points in a subgrid is less than 80% of the average data point density of the region, it is marked as a data sparse area. The current average density is set to 3000 data points per square kilometer, so a subgrid with less than 2400 data points per square kilometer is marked as a sparse area. For these areas, the system performs interpolation operations. The interpolation algorithm uses the inverse distance weighting method, and the target point is influenced by its surrounding up to 16 nearest known points. The weight of each point is equal to the square of the inverse of the distance from the target point. For example, if the nearest points are 10 meters, 20 meters, and 30 meters away from the target point, their weights are 0.01, 0.0025, and 0.0011, respectively. The final interpolation result is the sum of all weights multiplied by the attribute value of the corresponding known point, divided by the total weight sum.

[0044] Subsequently, the system performs spatial interpolation mapping for all subgrids to form a preliminary geological information distribution map. This map maps the representative attribute value of each subgrid in space and connects them to generate a continuous distribution image. To ensure image smoothness, the system detects all abnormal values in the subgrids. The detection standard is that if the attribute value of a subgrid is more than 2 times the standard deviation of all grid attributes in the region from the average value of its 8 adjacent grids, it is marked as an abnormal point. For example, if the standard deviation is 5 and the average value is 30, grids with attribute values exceeding 40 or below 20 are considered abnormal points. The abnormal point processing method is smooth replacement, i.e., replacing its value with the average attribute value of the adjacent 8 grids.

[0045] Finally, to form the final geological information synthesis dataset, the system performs multi-layer grid information fusion. The fusion process starts from the bottom subgrid and integrates upwards step by step. In each fusion level, the final class of the upper grid depends on the majority class of all subgrids below it, and if there are equal numbers of classes, the class with more sample points is selected. For example, a certain upper grid has 4 subgrids under it, with 2 of class A and 2 of class B, and the total number of A class points is 200 and the B class is 150. Select A as the representative class of the upper grid.

[0046] S2 includes adopting a data processing tool to hierarchically analyze the sub-region set for the geological information and the preliminary classification result, to obtain the distribution characteristics of the abnormal value points and boundary information, and to determine a preliminary abnormal distribution range; according to the abnormal distribution range, in combination with a depth analysis means, local feature comparison is performed on the abnormal value points and boundary information to obtain an abnormal region distribution map; through the distribution map, a preset threshold is used to perform screening processing on the abnormal region to obtain a candidate point set of key features; in combination with the distribution law of potential features, the support vector machine algorithm is used to classify and optimize the point set for the candidate point set of key features; if the key geological feature points are unevenly distributed in the sub-region set, the point set is adjusted through a data smoothing tool to obtain an evenly distributed feature point set; according to the evenly distributed feature point set, in combination with the spatial characteristics of the geological information, a spatial mapping technology is used to associate and integrate the point set to obtain a comprehensive geological feature distribution map; through the comprehensive geological feature distribution map, in combination with the intermediate result in the screening process, the potential features are verified again to determine the final key geological feature point set.

[0047] In this embodiment, first, the preliminary classification result is integrated with the original geological information to form a feature data set containing 12 dimensions of spatial coordinates, lithology identification, structural plane orientation, fracture development index, weathering degree, porosity, etc. For each sub-region, the system uses a three-dimensional cubic sliding window with a fixed side length of 30 meters to move the window step by step in the entire region, and the data points in each position within the window are counted. The statistical indicators include the average value and the standard deviation of the lithology code, which are used as the basis for judging abnormal points. For each data point within the window, the system calculates the difference between its lithology value and the average value of the window. If the difference is greater than 2 times the standard deviation of the window, the data point is determined to be an abnormal point. The 2 times standard deviation is the default abnormal judgment threshold of the system, which is derived from the statistical analysis results of the lithology fluctuation range of more than 100 different structural regions, and has universality and can effectively control the false positive rate. The side length of the sliding window is 30 meters, which is determined according to the spatial resolution of the unmanned aerial vehicle three-dimensional modeling and the minimum structural partition size in field exploration, which meets the local feature extraction accuracy requirements and also considers the calculation efficiency.

[0048] Meanwhile, the system traverses the classification results of adjacent sub-regions, and if the lithology category labels of two adjacent sub-regions are different and the spatial distance between them is less than 20 meters, the interface between them is identified as a geological boundary line. The spatial threshold of 20 meters is set according to the commonly used structural belt identification interval in geological surveys, to ensure the sensitivity and accuracy of boundary identification. After the identification of abnormal points and boundaries, the system maps all abnormal points to a three-dimensional model according to spatial coordinates, and analyzes these abnormal points using a density-based clustering algorithm. The clustering algorithm takes each abnormal point as the center, sets the clustering radius to 30 meters, and counts the number of abnormal points within the radius. If the number is not less than 5, it is considered that the point is located in an abnormal cluster area. The 30-meter clustering radius is consistent with the length of the sliding window described above, facilitating spatial consistency processing, and the 5-point clustering judgment threshold is obtained by analyzing the minimum density of abnormal points in common fault structures.

[0049] For each abnormal cluster area, the system calculates the average attribute difference between the cluster center point and the edge point in the dimensions of lithology, structural surface dip angle, and weathering degree. If the difference in any dimension is greater than 20%, the cluster area is retained as a candidate key area. The 20% difference ratio is determined through manual exploration and verification of multiple typical structural areas, which can significantly distinguish structural anomalies from general geological disturbances. All abnormal points in the candidate area are extracted as a candidate point set and input into the trained support vector machine model for binary classification prediction. The model is a radial basis function kernel model trained using 3000 labeled samples, with a 12-dimensional feature vector as input and a prediction label indicating whether the point belongs to a key geological feature point as output.

[0050] Points with key labels output by the model are collected as a key point set. Next, the system performs spatial balance analysis on the key point set. The specific steps are to divide the entire study area into 1 square kilometer grids and count the number of key points in each grid. If the ratio of the maximum value to the minimum value exceeds 3 times, the system considers that the point distribution is uneven. The 3 times density ratio threshold is an empirical threshold derived from the minimum constraint standard for point sampling balance in multiple geological mapping projects nationwide. If it is determined that the distribution is uneven, perform data smoothing: for areas with a point density greater than the average value, remove some points in proportion, aiming to keep the density of these areas not more than 120% of the average value of the entire area; for areas with a point density less than the average value, the system adds new smoothed points by performing linear interpolation between each point and its eight neighboring points, until the density of the area reaches more than 80% of the average value, ensuring the stability of the point density.

[0051] After the point density adjustment, the system constructs a connection graph between points in three-dimensional space. The system takes each point as a node and calculates the spatial distance and attribute similarity between it and other key points within 50 meters. If the spatial distance between two points is less than 50 meters and the lithology code is completely consistent, an edge relationship is established. The 50-meter connection threshold is derived from the average extension length of common fractures or contact zones in actual field geological structures. All connected points form a structural connected subgraph, and the system performs a depth-first traversal on each subgraph to determine whether it meets the following three indicators: First, the spatial connection integrity between all points is not less than 95%, i.e., on average, each point is connected to at least 5 other points; second, the attribute difference value is less than 10%, i.e., the lithology difference between any two connected points is not more than 10% of the lithology code in the entire dataset; third, the overall spatial trend consistency is greater than 90%, i.e., the point extension direction and the main structural direction of the region have an angle less than 10 degrees. These three judgment criteria are derived from the system's logical deduction of geological models and the calculation of a large number of measured structural line directions and point attributes.

[0052] Any connected structural subgraph that meets the above three standards will include a point set that will be the final set of key geological feature points, output for subsequent path planning, sample selection, and geological interpretation application modules. This process ensures that all output points are real abnormal clusters, have spatial consistency, structural continuity, and attribute stability, and are high-confidence geological key points with repeatability, verifiability, and engineering applicability.

[0053] The candidate point set is classified and optimized using a support vector machine algorithm to accurately identify and screen key geological feature points. The steps are as follows: First, extract feature parameters for each point in the candidate point set, including but not limited to spatial coordinates (such as latitude and longitude or three-dimensional position), geological attribute values corresponding to the point (such as rock type, stratum thickness, mineral content, color information, reflectivity, etc.), and local variation gradient values with surrounding points. A multi-dimensional feature vector is constructed for each point, where the dimension values are normalized to scale all input features between 0 and 1 to eliminate the bias caused by different dimensions on model training. Then, according to the existing part of the artificial labeling of key feature points and non-key points data to construct the training set, the support vector machine classifier is trained by supervised learning. During training, the radial basis function kernel is selected to map the input feature space to a high-dimensional feature space to enhance the separability, and the kernel function parameters such as the penalty factor are set to 100, and the kernel width parameter is set to 0.5. The value is selected from multiple candidate values by 5-fold cross-validation method to obtain the optimal combination corresponding to the minimum error. The objective function is to minimize the weighted sum of classification error and interval maximization, and the optimal hyperplane is obtained by using the Lagrange multiplier method. After training, the feature vector of all candidate points is predicted using the classification model, and the model output result is the classification label of each point, i.e. key point or non-key point, with the confidence score of each classification. The system sets the confidence threshold to 0.8, which is an empirical value and is based on the standard of ensuring that the classification accuracy is greater than 90% in multiple validation processes. Only the points with a classification result of key points and a confidence score greater than or equal to 0.8 are retained as the final key geological feature points. After classification, the system performs spatial distribution statistics on all key points to analyze their number distribution in each sub-region. If it is found that the number of key points in some sub-regions is significantly lower than 30% of the average value, it is considered that the feature points are unevenly distributed in the region, and the subsequent data smoothing step is entered. This calculation process ensures that the support vector machine has the ability of supervised learning, and realizes high-precision classification and optimization of complex geological feature points in a high-dimensional feature space, so as to accurately identify key points with geological significance and provide high-quality basic data for subsequent spatial mapping and feature verification.

[0054] S3 includes constructing a feature correlation framework for the correspondence between the geological feature point set and the environmental parameters through data integration means, obtaining preliminary correspondence degree distribution data; according to the preliminary correspondence degree distribution data, a preset threshold is used for screening, if the correspondence degree is lower than the preset threshold, the point set is corrected to obtain the corrected point distribution data set; according to the corrected point distribution data set, through the data comparison tool, the matching deviation between the point set and the environmental parameters is analyzed, and the deviation point subset is determined; according to the point subset, a spatial adjustment method is used to calibrate the position of the point subset, and an adjusted point distribution structure is obtained; according to the adjusted point distribution structure, through the feature correlation framework, the correspondence degree between the point set and the environmental parameters is recalculated, and updated matching distribution data is obtained; according to the updated matching distribution data, if the matching distribution data still has points that do not reach the preset threshold, these points are optimized again through the data smoothing tool, and a final optimized distribution map is obtained; according to the final optimized distribution map, the point set and the geological features are integrated by using the spatial mapping technology, and a comprehensive geological feature distribution view is determined.

[0055] In the embodiment, for the correspondence between the geological feature point set and the environmental parameters, a multi-dimensional attribute matrix is first established, wherein the rows represent each geological feature point and the columns represent the attributes and environmental parameter values. The point attributes include 12 items such as lithology code, structural surface inclination, fracture density, weathering index, porosity, color classification, hydrological type, surface sedimentation type, fault zone distance, historical activity level, etc., each attribute is obtained by unmanned aerial vehicle remote sensing and manual measurement data, the units are consistent, and the numerical values are standardized; the environmental parameters include 10 items such as stratum thickness, regional slope, vegetation coverage, groundwater depth, historical earthquake density, annual precipitation, surface temperature, rock layer resistivity, surface erosion intensity, and surrounding structural complexity, which are all from professional databases or remote sensing inversion results. By constructing a feature correlation framework, 12 attributes of each geological point are taken as independent variables, and 10 environmental parameters are taken as dependent variables. A local regression model is constructed by using point-by-point multivariate linear regression method. First, the data of each point in the geological feature point set is extracted, and each point is taken as an independent regression unit. The input variables of each point include 12 attributes such as lithology code, structural surface inclination, fracture density, weathering level, porosity, hydrological type code, color classification value, fault distance, fault intensity index, stratum thickness code, structural stability index, and lithology hardness score. The corresponding output variable is the environmental parameter value of the spatial position of the point, including 10 items such as the slope value, groundwater depth, vegetation coverage, annual precipitation, earthquake activity frequency, resistivity, erosion intensity, sedimentation type number, surface temperature, and structural disturbance level.

[0056] Before performing the point-by-point regression, the system normalizes all variables to make their values fall between 0 and 1, so as to eliminate the influence of dimension and improve the stability of the regression model. Then, for each point, the attribute values and environmental parameter values of the point and its surrounding neighboring points within the range of the sub-region where the point is located are extracted to form a local sample set. The size of the local sample set is determined by the spatial density of the neighboring points, and the minimum sample number is set to 20 and the maximum sample number is set to 100. If there are not enough sample points collected within a range of 10 meters, the range is expanded to 20 meters until the minimum sample number is met.

[0057] Each local sample set is used to establish a multiple linear regression model for the point. In the model, 12 attributes are used as independent variables, and each of the 10 environmental parameters is used as a dependent variable, and 10 regression equations are constructed respectively. The least squares method is used to fit each equation. The system calculates the coefficient of determination R square value for each regression equation, which is a measure of the ability of the regression equation to explain the variance of the dependent variable. The system takes the average of the 10 R square values as the overall "matching degree score" of the point, which is used to measure the overall matching degree of the point attributes and environmental parameters. If it is found during modeling that the standard error of a variable is greater than 50% of its regression coefficient, it means that the variable contributes insufficiently to the regression result, and the system will automatically exclude the variable and rebuild the model to ensure the effectiveness and stability of the model.

[0058] After modeling is completed, the matching degree scores of all points are saved and used in subsequent matching screening, correction and optimization analysis process. The regression analysis is performed independently for each point, with strong local adaptability and high data interpretation accuracy, especially suitable for three-dimensional geological modeling scenarios with strong geological environment heterogeneity and complex data structure. The whole process is automatically processed without human intervention, ensuring the objectivity and repeatability of the calculation process, and meeting the high requirements of geological information accuracy and environmental consistency in actual engineering.

[0059] The coefficient of determination R square value calculated by each model is used as the "matching degree score" of the point. The R square value represents the ability of the attribute value to explain the change of the environmental parameter, and the value range is 0 to 1. The larger the value, the higher the matching degree. The matching degree scores of all points are converted to percentage to form the preliminary correspondence degree distribution data.

[0060] According to experience and a large number of historical exploration data analysis, when the R square value is higher than 0.8, the correlation of geological points and environmental parameters has statistical significance, so 80% is set as the screening threshold. In the screening process, the point corresponding to the degree lower than 80% is identified as a point to be corrected. The correction process first locates the point in the spatial 10-meter radius range of the adjacent point set, calculates the attribute mean μ and standard deviation σ of the set, and then compares the difference Δ of the current point and μ. If Δ is greater than 50% of σ, the attribute value is corrected by 50% of Δ towards μ, and all attributes are processed one by one. After correction, the new matching degree score is calculated by re-executing the multiple regression, and the point data is updated. All corrected points are recombined to form the corrected point distribution data set.

[0061] Then, error analysis is performed on the corrected points. The system calculates the difference between the attribute value of the point and the mean value of the grid unit where the point is located. If the difference is greater than 20% of the standard deviation of the environmental parameter, the point is classified as a deviation point. The deviation point subset is input to perform spatial adjustment. Specifically, the main structure direction θ of the local area where the point is located is determined, and the linear trend direction is fitted by constructing a spatial vector based on the coordinates of the surrounding 8 points. If the angle between the direction and the gradient direction of the environmental parameter is less than 30 degrees, it is determined as an adjustable area. The point is offset by 50% of the original error value along the direction, and the offset is measured in meters. After adjustment, a new point spatial coordinate is formed.

[0062] The matching degree of all the adjusted points is calculated again to generate the updated matching degree distribution data. Points with a matching degree still lower than 80% enter the smoothing optimization link. The smoothing operation uses the inverse distance weighted interpolation method to average the attribute values of the 8 points around the point according to the distance weight. The closer the distance, the higher the weight. The weight calculation method is the inverse of the distance between the target point and the neighboring points, and all weights are normalized. The interpolation value replaces the original point attribute value. After interpolation, the matching degree is calculated again to form the final optimized point set.

[0063] Finally, all points are remapped to the three-dimensional geological model to construct a comprehensive geological feature distribution view. In the mapping process, each point is assigned to the corresponding voxel unit according to its three-dimensional coordinate position. If a voxel contains multiple points, the attribute value of the point with the highest matching degree is used as the final value. The comprehensive view supports layer-by-layer profile display, full-space query, and environmental factor visualization linkage, and is used as part of the geological mapping output results for geological structure analysis, mineral prediction, and engineering site selection.

[0064] S4 includes adopting a spatial analysis tool to compare the sampling point position set with the distribution map of geological features, judging the coverage proportion of the sampling point position in the high-density area, if the coverage proportion does not reach the preset threshold, adjusting the position of the sampling point, obtaining the adjusted sampling point distribution data set; according to the adjusted sampling point distribution data set, calculating the relevance of the sampling point position and the feature distribution, analyzing whether the point position distribution in the high-density area is balanced, if the distribution deviation exceeds the preset range, locally optimizing the point position, obtaining the optimized point position distribution structure; for the optimized point position distribution structure, adopting a spatial interpolation method to process the feature distribution in the grid division, obtaining the regional analysis data, determining the boundary range of the density area; according to the boundary range of the density area, through a data screening tool, secondarily checking the sampling point position, judging whether the point position falls within the boundary range, if there is a boundary deviating point position, adjusting the position of the deviating boundary point position, obtaining the corrected sampling point set; for the corrected sampling point set, adopting a data integration framework, associating and mapping the position data with the distribution characteristics of geological features, obtaining a comprehensive sampling point distribution view; according to the comprehensive sampling point distribution view, through a visualization processing tool, superimposing and comparing the sampling layout and the regional analysis result, determining the final sampling point position configuration scheme; for the final sampling point position configuration scheme, adopting a data storage module to archive the position data and the feature distribution information, obtaining a structured sampling layout data set.

[0065] In the embodiment, the system first performs spatial superposition on the input initial sampling point position set and the geological feature distribution map, and analyzes whether each sampling point is located in a key geological high-density area. The judgment basis of the high-density area is whether the number of geological feature points in a unit area is greater than 150% of the average point density of the whole area. For example, there are 30,000 key points in a 200 square kilometer range, and the average point density is 150 points per square kilometer, so the threshold of 150% is 225 points per square kilometer. The system constructs a buffer zone with a radius of 20 meters centered on each sampling point, and counts whether the area in the buffer zone contains an area with a density exceeding the threshold, and if it does, the point is determined to cover the high-density area. The system counts the proportion of the number of sampling points located in the high-density area to the total number of sampling points, and if the proportion is less than 80%, the sampling point position is adjusted. 80% as the lower limit of the coverage rate is an empirical value obtained through the sample balance analysis of nearly 50 actual exploration areas, which can balance the sampling representativeness and point position cost control. For the point position that needs to be adjusted, the system calculates the shortest spatial distance from the point to the nearest high-density boundary, and moves the point along the direction by 70% of the distance, and records the new position as the updated sampling coordinates. The 70% offset coefficient is determined according to the spatial decay model of the point density transition area, which can ensure that the point position is close to the key area without excessive concentration.

[0066] According to the adjusted sample point set, the system divides the high-density area into equal-area grids, usually set to 10 meters by 10 meters per grid. For each grid, the number of sample points in it is counted, and the maximum, minimum and average values are recorded. Calculate the distribution deviation ratio, which is the difference between the maximum and minimum values divided by the average value. If the ratio exceeds 30%, it is judged that the point distribution is uneven. The 30% threshold is determined by the error trend of the identification accuracy of different sampling densities on the geological structure in the simulation experiment, and exceeding this threshold will significantly reduce the profile connectivity. The system adjusts the points in the high-density grid area where the number of points exceeds the standard. Each time the points are moved no more than 10 meters towards the center of the adjacent low-density area, with priority given to the direction with the weakest connectivity, while ensuring that they do not cross the boundary into the low-density area.

[0067] The optimized sampling point data is sent to the spatial interpolation module, and the inverse distance weighted interpolation method is used to process the attribute values in each grid. The minimum influence point is set to 6, the maximum influence point is set to 12, the weight is the square reciprocal of the distance, and the interpolation radius is set to 20 meters. Each target grid obtains an estimated value by weighted average of the attributes of the adjacent known points, and finally forms a continuous geological attribute interpolation layer. The system analyzes the layer and extracts the contour lines of the geological feature density. The system sets the contour extraction threshold to 130% of the average density, that is, if the average density is 150 points per square kilometer, the area with a density of more than 195 points is defined as the feature boundary. The boundary extraction uses a standard contour construction algorithm to connect points with the same density value to form a continuous closed figure. The specific calculation process is as follows: First, from the geological attribute layer generated by the inverse distance weighted interpolation, the system traverses all grid cells according to the regular grid structure, extracts the interpolation result value of each grid, and the value represents the density value of the geological feature point at that position. The system calculates the average density of all interpolation values in the layer, which is recorded as the base density value. Assuming that the average density is 150 points per square kilometer, the set contour extraction threshold is 130% of the average value, that is, 195 points per square kilometer. Then, in the traversal process, the system identifies all grid boundaries that cross the critical value of the equal density (i.e. 195 points per square kilometer). The method is as follows: for any two adjacent grid cells, if one of the interpolation result values is higher than 195 and the other is lower than 195, it is considered that the boundary is crossed by the contour line. The system inserts an intermediate point into the boundary, and the calculation method is linear interpolation. Specifically, if the two adjacent grid points are P1 and P2, and the corresponding density values are D1 and D2, respectively, then the equal point P is calculated by linearly weighting the coordinates between P1 and P2 according to the equal density point. That is, when 195 is between D1 and D2, the relative position of the P point on the line segment P1P2 is calculated so that the density value is 195. After completing the interpolation point calculation of all crossing boundaries, the system connects all such interpolation points into line segments according to the grid topology. Each line segment represents the direction of the contour line between the grids. The system further uses a topology tracking algorithm to construct the contour line path. This algorithm finds each possible crossing state by grid, and connects multiple boundary line segments into a complete path. The system checks the connectivity of each path to see if it forms a closed region, and if it forms a closed loop, it is marked as a complete contour figure. Then, the system performs object-oriented structure recognition on all contour closed figures, matches each closed figure with its contained interpolation point density value, and filters out abnormal lines with an area less than 10 square grids or an internal density average value less than 195, and retains closed regions with an internal density greater than or equal to the contour threshold as valid geological feature density boundaries. Finally, the system outputs the closed figures composed of these contour lines as a vector layer, which can be used for subsequent boundary partitioning, path avoidance, or key area planning analysis.The calculation logic of this standard contour construction algorithm can ensure the continuity, accuracy and closure of the boundary extraction result. The system then performs a spatial intersection analysis of all sample points with the boundary layer. If a sample point is more than 5 meters away from the nearest boundary, it is marked as a deviation point. The 5-meter distance threshold is based on the modeling error of the unmanned aerial vehicle and the actual surveying error limit, which can ensure the data implementability. For deviation points, the system translates them to the inside of the boundary line by 3 meters in the direction of the boundary, updates their coordinates, and inputs the corrected sample point set into the data integration framework. The corrected sample point set is bound with the geological attribute data to form the final spatial attribute structured record. Finally, the system displays the above integrated data and the geological area analysis layer through the visualization module, and marks different types of high-density coverage points, corrected points, boundary inside and outside points, etc. on the map with different colors to provide an intuitive layout result. The system scores based on three indicators: final point coverage, uniformity, and boundary fitting. The scoring results are stored in the database, along with the final sample point coordinates, attribute fields, and grid numbers, etc. as structured data sets, which can be exported in CSV, GDB or GeoJSON format for subsequent sampling execution system or analysis software to call.

[0068] S5 includes, for the preliminary optimized navigation path scheme, using a data processing tool to analyze the matching degree of the path scheme and the terrain, and if the matching degree is lower than a preset threshold, locally correcting the path scheme to obtain a corrected path distribution; according to the corrected path distribution, calculating the correlation between the sample points and the navigation path through a spatial analysis module, and if the correlation cannot cover the target area, redistributing the point distribution to determine an adjusted point layout; for the adjusted point layout, using an information integration framework to fuse the path scheme and the passing cost data to obtain a comprehensive path cost distribution view; according to the comprehensive path cost distribution view, identifying the cost segments in the navigation path that are higher than a preset threshold through a data filtering tool, and if the proportion of the cost segments that are higher than the preset threshold exceeds a preset proportion range, planning a replacement path for the cost segments that are higher than the preset threshold to obtain an optimized path structure; for the optimized path structure, using a spatial interpolation method to analyze the coverage range of the path scheme and the sample points to obtain regional distribution data; according to the regional distribution data, evaluating the compliance of the path scheme with the planning requirements, and if the compliance does not meet a preset standard, fine-tuning the path structure to determine a final navigation path configuration.

[0069] In this embodiment, the system first performs a terrain matching analysis on the input preliminary optimized navigation path scheme. The system reads the three-dimensional coordinate sequence of the path scheme, segments it into consecutive 10-meter path unit segments, and extracts the three-dimensional direction vector of each segment. At the same time, the system extracts the normal vector of the corresponding ground surface unit under the path segment from the digital elevation model, and calculates the angle θ between the direction of each path segment and its corresponding terrain normal using the vector angle calculation formula. The average angle value of all path segments is calculated, and if the average angle is greater than 15 degrees, the path is determined to not meet the terrain adaptation requirements. 15 degrees is the maximum adaptive slope angle determined by the actual passing capacity of personnel and equipment in geological field work, and exceeding this value will result in a sharp increase in the risk of passing. The system marks all path segments with an angle greater than 20 degrees as high-risk segments, inserts control nodes before and after the segment, and reconstructs the path curve using a cubic Bezier interpolation method, while setting a curvature constraint to ensure that the turning radius of each segment is not less than 20 meters, ensuring that the path remains continuous and smooth in high-slope areas without sharp angles or turning problems.

[0070] After completing the local correction of the path, the system loads the current set of sampling points and constructs a spatial buffer zone with a radius of 50 meters centered on the navigation path, and detects whether each point falls within the buffer zone. The number of covered sampling points is calculated, and if the coverage rate is less than 80%, it is marked as insufficient coverage. The 80% coverage threshold is based on the relationship between sampling efficiency and operation efficiency in 50 typical mapping area projects, and a coverage rate below this value will significantly increase the probability of overlapping and missing sampling. For points that are not covered, the system calculates the Euclidean distance from the point to the nearest point on the path, and performs a shortest path projection within a range of not more than 30 meters to generate new point coordinates, ensuring that the new point is located in the path accessible area. All adjusted point positions are updated to form the adjusted point layout.

[0071] Subsequently, the system combines the path coordinates and surface feature data to model the path passing cost. The passing cost of each path segment is calculated by weighting and superimposing three parts: the first part is the terrain slope factor, the slope value ranges from 0 to 45 degrees, and the system divides it into five levels with weights of 0.1, 0.2, 0.3, 0.4 and 0.5; the second part is the vegetation density factor, which is obtained by the NDVI value in the remote sensing image, divided into four levels with weights of 0.1, 0.2, 0.3 and 0.4; the third part is the passing time factor, which is assigned a weight of 0.3 if it is during peak hours, and a weight of 0.1 if it is during normal hours. The cost value of each path segment is the weighted sum of the above three factors, and the system generates a passing cost heat map for all path segments and calculates the proportion of path segments with a cost value exceeding 0.7 to the total path length. If the proportion exceeds 20%, it means that there are too many high-cost segments. 0.7 is the upper limit warning value for passing safety, and 20% is the upper limit of the acceptable risk segment proportion, both of which are based on the stability evaluation of unmanned equipment passing and historical operation data standards.

[0072] For all high-cost segments, the system creates a path replacement search grid within 30 meters of its path segment, performs a cost-minimized path search using the A-star algorithm, and corrects path connectivity using the Dijkstra algorithm. The replacement path must meet the following criteria: the same start and end points, a path length not exceeding 1.2 times the original path, and an average segment cost of less than 0.5. After the replacement path is complete, the system replaces the original high-cost segment with the optimized path structure. In this path optimization process, the system first identifies all high-cost path segments and generates and replaces the replacement path according to the following steps: First, the system performs spatial analysis on each high-cost path segment, calculates the center line, and expands 15 meters on both sides of the center line to form a path search area with a total width of 30 meters. The area is divided into a regular grid with a grid length of 5 meters, forming a search grid matrix. Each grid cell is assigned a travel cost value, which is calculated based on multiple source data, including terrain slope, surface material type, water body distribution, and presence of obstacles. Each factor is assigned a standard weight, with a slope weight of 0.3, a surface material weight of 0.3, a water body obstacle weight of 0.2, and other obstacle weights of 0.2. The weighted calculation results in the travel cost of each grid. Next, the system sets the start and end points of the original high-cost path segment as the fixed start and end points of the replacement path, and calls the A-star path search algorithm in the constructed grid. The A-star algorithm uses travel cost as the cost function and sets the heuristic function as the Euclidean distance from the start point to the end point, preferentially searching for grid cells with shorter expected path lengths and lower costs. The system calculates the sum of the current path cumulative cost and the expected target path estimate in real time during the search process, and selects the grid with the smallest value as the expansion point. The search result is a continuous grid path that meets the cost minimization criteria. The path result generated by the A-star algorithm is then corrected for connectivity using the Dijkstra algorithm to ensure that there are no gaps between path segments. The Dijkstra algorithm starts from the path start point and re-evaluates the minimum cumulative cost from the start point to each node based on the existing path until the end point is confirmed to be optimally connected, solving the local optimal trap problem caused by the heuristic function and ensuring the entire path segment is smooth. After the path is generated, the system performs conditional verification on the replacement path, with the following criteria: the total path length must not exceed 1.2 times the original high-cost path segment; the average travel cost of each grid cell in the replacement path must be less than 0.5; and the start and end points must coincide with the original path. If all the above conditions are met, the system deletes the original high-cost path segment and inserts the replacement path into the original path structure. The replaced path is then re-labeled and participates in the overall optimization evaluation process of the next round of navigation path. Through this calculation process, the system can accurately construct a replaceable path in high-travel-cost areas, effectively reduce the overall path cost, and improve the path's passability and task execution efficiency.

[0073] Next, the system performs the area coverage analysis of the path and the sampling points again. The system divides the entire target area into cells, each with a size of 25 meters by 25 meters, and counts the number of sampling points in each cell and whether it is within the path buffer. Using the inverse distance weighted interpolation method, the coverage density of the points within the path buffer is estimated. If there is a continuous area with a total area of more than 10% but no sampling points falling into it, it is considered a blind area. 10% is the minimum control lower limit of spatial coverage evaluation, and exceeding it will significantly affect the representativeness of the data. The system will add auxiliary path nodes in the direction of the blind area and update the path structure so that all areas are covered.

[0074] Finally, the system verifies the compliance between the final path structure and the planning requirements through the data verification module. The evaluation indicators include: whether the average path cost is less than 0.5; whether the proportion of high-cost sections is less than 20%; whether the blind area is less than 10%; and whether the proportion of path buffer coverage sampling points is greater than 85%. All four indicators must be met. If not, the system starts the path structure fine-tuning mechanism to improve the indicators to the target range by fine-tuning node coordinates, adjusting buffer directions, optimizing turning paths, etc. After passing the verification, the system encapsulates all the control point coordinates, section cost values, buffer coverage results, and corresponding point association numbers of the path into a structured data set and stores it in a standardized JSON format and GDB geographic database format for direct calling in subsequent path execution, navigation scheduling, sample arrangement, etc.

[0075] S6 includes obtaining dynamic environment information from the field investigation scene, continuously monitoring environmental changes, obtaining the latest terrain and obstacle data distribution; based on the terrain and obstacle data distribution, using the pre-established path update rules, comparing and analyzing the navigation path information, if the terrain obstacles affect the path passage are detected, triggering the path adjustment process, determining the preliminary adjusted path scheme; for the preliminary adjusted path scheme, using spatial analysis tools to calculate the matching degree of path information and dynamic environment data, if the matching degree is lower than the preset threshold, modifying the local section of the path, obtaining the modified path layout; based on the modified path layout, using the information fusion framework, integrating and processing the real-time collected environmental change data and path information, obtaining the comprehensive environmental adaptability distribution view; for the comprehensive environmental adaptability distribution view, identifying the paragraphs in the navigation path affected by the terrain obstacles, if the proportion of the identified affected paragraphs exceeds the preset range, starting the re-planning process, determining the alternative path scheme; based on the alternative path scheme, using the spatial interpolation method to calculate the path layout and the coverage range of the field investigation scene, obtaining the path distribution data; for the path distribution data, evaluating the fit degree of the path scheme and the investigation scene demand, if the fit degree does not reach the preset standard, adjusting the path layout, determining the final navigation path configuration.

[0076] In this embodiment, the system first continuously collects data including terrain elevation changes, three-dimensional obstacle distribution, vegetation coverage, surface albedo, real-time wind speed and pressure changes, etc. every 5 seconds using high-frequency environmental perception units deployed in the field investigation area and laser radar equipment carried on the unmanned aerial vehicle. All data are synchronously mapped into a unified three-dimensional coordinate system based on a geographic reference system and stored in a structured manner with a grid resolution of 5 meters by 5 meters. The system calls the path monitoring module to perform registration processing on the current navigation path and the batch of dynamic environmental data. The registration method is to perform Euclidean distance calculation on the center coordinates of each segment of the path and the current obstacle data set. If the shortest distance of any path segment to the obstacle is less than 5 meters, or the slope value of the current segment in the elevation grid increases by more than 20% from the initial value, it is considered that the path segment is impassable due to obstacles or terrain changes. The 5-meter distance threshold is the minimum avoidance space for unmanned ground equipment, and the 20% slope mutation is the acceptable change upper limit based on human traffic stability experiments.

[0077] For all path segments determined to be impassable, the system initiates a path adjustment process. The specific operation is to expand 10 meters in front and behind the blocked segment based on its start and end points to form a total length of 20 meters adjustment section, and insert 3 new control nodes inside the section, with a fixed node spacing of 3 meters. The insertion rule is to determine the node offset direction according to the maximum density direction of the obstacle distribution. The system uses a cubic Bezier interpolation algorithm to replace the original path segment with a new smooth curve segment, ensuring that each control point is at least 5 meters away from the nearest obstacle, and the interpolation result satisfies the turning curvature radius of not less than 10 meters, ensuring the safety of passing and the stability of operation. When adjusting the path segment determined to be impassable, the system performs the following curve reconstruction process based on cubic Bezier interpolation to ensure that the new path curve avoids obstacles and meets the turning safety constraints: First, the system determines the start and end points of the impassable path segment, and extends 10 meters in front and behind the path direction to form a 20-meter-long path adjustment section. This adjustment section is used to insert a new path curve. The system analyzes the obstacle distribution in the adjustment section, calculates the gradient vector of the obstacle density in each direction, and identifies the direction of obstacle concentration. Then, according to the maximum density direction of the obstacle distribution, the lateral offset direction of the control node is set, that is, the control point is away from the most dense area of the obstacle, and the offset direction is perpendicular to the obstacle gradient direction. The system then inserts 3 new control nodes uniformly inside the path adjustment section, with a longitudinal spacing of 3 meters. The offset distance of the control node in the direction perpendicular to the path axis is set according to the local obstacle density, with a maximum offset value of not more than 5 meters. Before each control point is inserted, the system calculates its Euclidean distance from the nearest obstacle in real time, and only when the distance is greater than 5 meters can the control point be confirmed; if it is less than 5 meters, it will be further offset in the opposite direction until it meets the 5-meter condition or cannot be offset. With the start point, 3 control points, and end point, a total of 5 control points, the system calls the cubic Bezier interpolation algorithm for path fitting. The cubic Bezier curve segment is determined by 4 control points. To generate a continuous and smooth curve segment, the system fits two segments: the first Bezier curve is defined by the start point and the 1st, 2nd, and 3rd control points, and the second curve is composed of the 2nd, 3rd control points and the end point, and is connected at the 2nd and 3rd control points through the tangent consistency and continuity conditions to form a whole curve with continuous first derivative. After the curve is constructed, the system uses differential geometry analysis method to calculate the local curvature radius of the path curve point by point, which is to calculate the ratio of the first and second derivatives at the curve interpolation point. All curvature radii must be greater than 10 meters. If there is a point less than this value, the system automatically adjusts the control point position or spacing until the minimum curvature radius of the entire path is greater than or equal to 10 meters. Finally, the system replaces the original impassable segment with the smooth path segment reconstructed by the Bezier curve, and adjusts the connection between the path segments to ensure the geometric continuity and passing stability of the entire path.This method effectively combines the obstacle avoidance requirements with the minimum turning radius limit of the navigation vehicle, ensuring that the adjusted path maintains drivability while achieving smooth and safe operation.

[0078] After the system completes the initial path correction, it enters the dynamic adaptability analysis phase. The analysis process involves spatially overlaying the corrected path layout with the latest environmental data. Using a rolling calculation method with a 10-meter sliding window, it evaluates parameters such as the number and density of obstacles, the fluctuation amplitude of local slopes, and the standard deviation of wind speed changes on a segment-by-segment basis. An environmental adaptability index is generated for each segment, ranging from 0 to 1, representing impassability to full adaptability. Sections with an obstacle density exceeding 3 per 100 square meters, a slope variation exceeding 25% of the initial value, and a wind speed variation exceeding 4 meters per second will be assigned an adaptability coefficient below 0.5. If the average adaptability index for all path segments falls below 0.7, or if the proportion of sections below 0.5 exceeds 10% of the total path length, the system will trigger the path replanning process.

[0079] During the replanning process, the system uses the endpoints of the impassable segment as the starting and ending points of the new route. It constructs a grid of candidate paths within a 30-meter radius, with each grid node spaced 5 meters apart. The system searches for the lowest cost path. The cost function increases the cost by 0.1 for every 5-degree increase in slope, 0.2 for every 1-meter decrease in obstacle distance, and 0.1 for every 1-meter-per-second increase in wind speed. The system selects the lowest-cost path as the new alternative and automatically seamlessly merges it with the original path.

[0080] After completing the splicing of alternative paths, the system regenerates a buffer zone based on the new path layout and performs a spatial analysis of the mission coverage of the entire path. This method establishes a 20-meter radius buffer zone centered on the path, divided into 10-meter grids. Each grid point is then determined to be covered by the path, and the total coverage ratio is calculated. If the coverage ratio is less than 90%, or if any single uncovered area exceeds 5% of the total mission area, the path solution fails to meet the mission requirements and the system performs fine-tuning operations. Specifically, the path node positions are nudged by no more than 2 meters at a time, with the direction determined by the principle of obstacle minimization, and local interpolation and smoothing are performed.

[0081] Finally, after meeting the three indicators of path safety, environmental adaptability and task coverage, the system integrates the three-dimensional coordinates of the path nodes, the adaptability index of each segment, the obstacle interference level, the path segment travel cost, the alternative segment mark and the fine-tuning record into a structured navigation data package, and generates a JSON format for field mission terminal equipment to call, and saves it in a standard geographic database format for subsequent operation scheduling, task adjustment and path tracking.

[0082] S7 comprises data for navigation path and sampling position, using a pre-established resource allocation model, analyzing the spatial distribution relationship between sampling position and path, obtaining a preliminary allocation scheme of resource allocation; according to the preliminary allocation scheme, the principle of balance is fused, the spatial analysis tool is used to evaluate the distribution uniformity of resource allocation, if the distribution uniformity is lower than the preset threshold, the resource allocation is locally adjusted, and the adjusted configuration layout is determined; for the adjusted configuration layout, a task allocation framework is constructed, a multi-target task is decomposed into a plurality of sub-task units, and a set of decomposed task units is obtained; according to the set of decomposed task units, a priority sorting method is used to calculate the matching degree of the sub-task unit and the resource allocation layout, and a task execution priority order is obtained; for the task execution priority order, the priority order is compared with the overall demand of the investigation scheme, if the comparison result shows that the task unit execution order is inconsistent with the overall demand, the priority order is adjusted, and an optimized execution sequence is determined; according to the optimized execution sequence, a data verification tool is used to analyze the fitting degree of the execution sequence and the complete execution scheme, and a final field investigation execution plan is obtained; for the final field investigation execution plan, a data mapping method is used to associate the execution plan with the actual distribution of navigation path and sampling position, and an execution path distribution is obtained.

[0083] In the embodiment, first, the navigation path and sampling point position data obtained from the three-dimensional modeling result of the unmanned aerial vehicle are analyzed, each sampling point is recorded in the form of three-dimensional coordinates, and pairing calculation is performed according to the spatial distance between the path segments. The pairing method used is to calculate the shortest vertical distance from each sampling point to the center line of each path segment. If the distance is less than or equal to 20 meters, the sampling point is attributed to the operation coverage range of the path segment. The basis for setting the 20-meter threshold is the line-of-sight control standard of the unmanned aerial vehicle and the results of multiple field flight coverage experiments. 20 meters is the boundary distance for stable control and effective information collection. After pairing is completed, a preliminary allocation model of resource allocation is constructed, that is, after the number of all sampling points distributed on the path segment is counted, the task resources (including operators, sample collection equipment, battery packs, etc.) are divided in proportion. The preliminary allocation strategy is that the number of resources for each path is equal to the number of sampling points multiplied by the single-point operation resource base. If the single-point resource base is set to 2 (1 person and 1 set of equipment), the path segment contains 4 sampling points, and the preliminary allocation resource is 8 units.

[0084] After the preliminary scheme is established, the system evaluates the balance of resource allocation. The specific method is to calculate the standard deviation of the resource quantity of each path segment, and then divide it by the average resource to obtain the coefficient of variation of resource allocation. If the coefficient of variation is greater than 0.25, it indicates that the resource allocation of each segment is quite different. The 0.25 threshold is derived from the statistical average upper limit of the resource usage fluctuation rate of 50 field tasks. At this time, the system will divide the excessive and insufficient resource segments into two sets, and adjust the excessive resource segments to the insufficient resource segments by the maximum-minimum difference adjustment method. The adjustment process moves by the smallest adjustment unit of 2 each time until the coefficient of variation is controlled within 0.25. This process ensures that all path segments obtain relatively balanced job resources and avoids resource redundancy or deficiency.

[0085] Next, the path segments and sampling point combinations after resource allocation are established into a task unit set. Each task unit contains the path segment number, its covered sampling point list, the allocated resource quantity, the task execution estimated time (the sampling point quantity multiplied by the average job time per point, and the average value is generally set to 15 minutes), and the geological anomaly intensity level of the area where the path segment is located. Then, the priority of each task unit is calculated. The calculation method is the weighted total score of four score items, including the geological intensity score (represented by 0.3, 0.6, 1 for low, medium, and high levels, respectively), the resource sufficiency score (the ratio of the actual allocation to the ideal value, and a value greater than 1 is assigned as 1), the path accessibility score (the reciprocal of the travel cost ratio, and the cost is calculated by combining the slope and obstacle statistics, with a range of 0 to 1), and the task urgency score (1 if the task scheduled completion time is less than 1 day from the current time, and 0.5 if it is more than 3 days). After standardizing the above scores, they are weighted and superimposed to calculate the priority score of each task unit.

[0086] After the system sorts the task units by priority, it compares the sorting result with the original task execution logic, i.e., checks whether it violates the task execution sequence dependency constraints, such as some sampling points must be performed after a specific path segment is completed. If the sorting does not meet this logic, the system automatically adjusts the task sequence topological sorting according to the dependency graph to generate a new task execution order that satisfies the dependency relationship. Then, the system uses the execution sequence verification module to calculate the logical fit degree of the new sequence, and checks the indicators including the logical consistency rate (the ratio of the sorting that meets the dependency logic), the resource use efficiency (resource unit utilization rate), and the path continuity (whether the path is connected according to the sorting execution). The weighted total score of the three scores is the fit degree total score. If it is less than 0.9, it returns to adjustment. The 0.9 threshold is the critical value of the stable scheduling scheme in the multiple rounds of task scheduling simulation tests.

[0087] The final determined task execution sequence is fused with the navigation path and the spatial position of the sampling points, specifically, each task unit is labeled with its execution sequence number, and its path segment and spatial position of the sampling points are uniformly mapped into the standard coordinate system map layer. The execution path data includes task sequence number, execution path segment number, covered sampling point coordinate list, required resource quantity, estimated execution time, priority score and other fields, and is output as a structured data table and a spatial vector graph file, which is called by the unmanned aerial vehicle scheduling system, the task scheduling system and the field operation system, to ensure that the task execution scheme is not only logically reasonable, but also spatially implementable, reasonably distributed in resources, and has traceability and verifiability.

[0088] S8 includes, for the field investigation execution scheme, preliminarily identifying potential risk points by pre-establishing a risk assessment framework, obtaining regional distribution information, and determining a preliminary risk distribution map; according to the preliminary risk distribution map, deeply analyzing the regional distribution information, if the analysis result shows that the risk value of the region exceeds the preset threshold, extracting specific risk points through a data screening tool to obtain a risk point set; for the risk point set, dynamically simulating the influence range of each risk point through a simulation analysis tool, obtaining boundary data of the influence range from the simulation result, and determining a risk influence range list; according to the risk influence range list, using a local adjustment mechanism to re-plan the part involving the risk points in the execution scheme, adjusting the related path and task allocation through a spatial distribution optimization tool to obtain an adjusted scheme segment; for the adjusted scheme segment, seamlessly connecting the adjusted scheme segment with other parts of the original execution scheme, if task allocation conflicts are found in the connection process, resolving the conflicts through a priority sorting tool to obtain an integrated complete scheme; according to the integrated complete scheme, using a verification tool to comprehensively detect the executability of the scheme, if the detection result shows that there are execution obstacles in the link, extracting an alternative path through a backup scheme database to determine a final investigation plan; for the final investigation plan, associating and matching the plan content with the actual environmental information of the field investigation to obtain environmental adaptability feedback, and obtaining an optimized execution guide.

[0089] In this embodiment, the system first parses the path segments, sampling points and operation time windows involved in the field investigation execution plan, and divides the plan area into standard grids with a side length of 50 meters as risk analysis units. Then the system calls the historical geological disaster database to extract the number of disasters in each grid, and divides the number by the maximum number of disasters in the study area to obtain the geological risk factor value. The slope risk factor is converted to a risk level by the average slope angle of each grid, set to 0.3 for less than 15 degrees, 0.6 for 15 to 30 degrees, and 1 for more than 30 degrees. The obstacle density factor is calculated by dividing the number of known obstacles in each grid by the area of the grid and normalized to a value between 0 and 1. The weather fluctuation factor is the proportion of the number of sudden weather events in the past 5 years to the total number of observations in the region. The passability factor is converted to the inverse form by the surface structure stability coefficient of the path through the grid, and the higher the score, the more difficult it is to pass. The system assigns weight values of 0.3, 0.2, 0.2, 0.2 and 0.1 to the above five factors, respectively. The weight setting is based on expert experience and past task evaluation regression analysis results. After weighted summation, the comprehensive risk value of each grid is obtained, and the high-risk grid is marked with a risk threshold of 0.7, which is set by fitting the past 80 field investigation accident data.

[0090] The system extracts all high-risk grids and marks the sampling points and path segments overlapping with them as a set of risk points. Then, the spatial simulation module is called to center on each risk point and conduct risk propagation analysis based on the gridded space propagation model. The propagation process expands outward in steps of 10 meters, and the risk gradient value is updated at each step until it is less than 0.05. The simulation boundary is the actual impact range of the risk point, and the final impact boundary list of all risk points is formed. According to the boundary, the system extracts the part of all path segments and sampling points in the task plan that falls within the range and forms a local high-risk task segment.

[0091] Next, the system performs local optimization adjustment on these high-risk task segments. First, the task segments are re-divided according to the path segments, and the sampling points are grouped based on whether they are affected. The path optimization uses the shortest path replacement strategy, which calls the risk avoidance path library to select a path that meets the shortest path and does not cross high-risk grids to replace the original path. When the sampling point is within the risk impact boundary, the point is moved outward to the center point of the nearest safe grid, and the moving distance is calculated by the spatial distance calculation module to find the lowest total passable cost among all possible moving paths. After adjusting the task segment, the system reassigns the operation resources and time windows. The operation resource calculation method is to configure 1 person and 1 device for each sampling point, and 2 people and 1 device for each path segment. The time window is calculated by multiplying the number of sampling points by the sampling time of 15 minutes per point, plus the path segment passable time, which is obtained by dividing the path length by the standard travel speed of 4 kilometers per hour.

[0092] Then, the system splices all the adjusted task segments with the unaffected task segments into a complete plan, with the requirement that the path segments are continuous and the time windows are conflict-free. If any time or resource overlap conflict is detected, the system reorders all the task segments according to their priority, which is the weighted sum of the geological importance (assigned a value from 1 to 3), the risk avoidance difficulty (assigned a value from 1 to 3), and the task urgency (arranged in descending order of the distance between the task deadline and the current time). If a segment conflicts with multiple tasks, the lower-priority task is executed later or the resource allocation is adjusted.

[0093] After the integrated complete plan is formed, the system calls the executability analysis module to check whether the path is continuous, the resources are sufficient and arrive on time, and the sampling order complies with the geological process logic. If any of the checks fails, the system calls the backup path database and the task alternative plan library for replacement until all the checks pass. Finally, the adjusted path segments, sampling points, operation times, and resource numbers are structured and summarized to generate the final survey plan.

[0094] Finally, the system compares the plan with the latest environmental data for adaptability, calls the topographic map, communication signal map, and weather forecast map layers, and matches each task unit in the plan one by one. If the match is not consistent, such as being in a signal blind area, a heavy rainfall area, or an impassable path area, the system calls the alternative segment for adjustment again until all the task points are in a safe and stable environment. Finally, a complete execution guide is formed, including the path map, the sampling task list, the resource allocation document, the risk avoidance explanation, and the environmental adaptation report, to ensure that the plan is operational, safe, and reasonably scheduled.

[0095] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements, and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A geological mapping method based on UAV 3D modeling, characterized in that: include: S1. By constructing a three-dimensional spatial data processing framework, an initial feature dataset is extracted from geological information, and the original data is decomposed into multiple sub-regional datasets to obtain a preliminary classified geological information set; S2. Based on the preliminary classified geological information set, analyze the outliers and boundary information in the sub-regional data set to determine the potential key geological feature point set; S3. For a set of key geological feature points, match the set of points with geological environment parameters. If the matching degree is lower than a preset threshold, recalibrate the points to obtain an optimized geological feature distribution map. S4. Based on the optimized geological feature distribution map, perform grid analysis on the feature distribution map to determine the initial sampling point location set; S5. For the initial set of sampling point locations, the travel cost under the terrain environment is taken into account. If the cost of the path segment is higher than a preset threshold, an alternative path is recalculated to obtain a preliminary optimized navigation path solution. S6. Based on the preliminary optimized navigation path plan, the real-time optimization mechanism is integrated to adjust the path. If terrain obstacles are detected, the path replanning process is triggered to determine the final navigation path. S7. Based on the final navigation path and sampling point location set, the multi-target survey task is decomposed into multiple subtask units, the task execution priority is obtained, and a field survey execution plan is obtained; S8. Conduct simulation analysis on potential risk points of the field investigation implementation plan. If the risk value exceeds the preset threshold, make partial adjustments to the plan and determine the final implementation plan. S6 includes acquiring dynamic environmental information from the field survey scene, continuously monitoring environmental changes, and obtaining the latest terrain and obstacle data distribution; based on the terrain and obstacle data distribution, using pre-established path update rules, comparing and analyzing the navigation path information; if a terrain obstacle is detected that affects the path, triggering a path adjustment process, and determining a preliminary adjusted path plan; For the initially adjusted path plan, the spatial analysis tool is used to calculate the degree of match between the path information and the dynamic environment data. If the match is lower than the preset threshold, the local segment of the path is corrected to obtain the corrected path layout. Based on the revised path layout, an information fusion framework is used to integrate the real-time collected environmental change data with path information to obtain a comprehensive environmental adaptability distribution view; Based on the comprehensive environmental adaptability distribution view, sections of the navigation path affected by terrain obstacles are identified. If the proportion of identified affected sections exceeds the preset range, a replanning process is initiated to determine an alternative path solution. Based on the alternative path solution, a spatial interpolation method is used to calculate the path layout and the coverage of the field survey scene to obtain path distribution data. Based on the path distribution data, the path solution is evaluated for its compatibility with the survey scene requirements. If the compatibility does not meet the preset standard, the path layout is adjusted to determine the final navigation path configuration. Said S8 includes executing a plan for field investigation, preliminarily identifying potential risk points by establishing a risk assessment framework in advance, obtaining regional distribution information, and determining a preliminary risk distribution map; Based on the preliminary risk distribution map, the regional distribution information is deeply analyzed. If the analysis results show that the risk value of the area exceeds the preset threshold, the specific risk points are extracted through the data screening tool to obtain the risk point set; For a set of risk points, a simulation analysis tool is used to dynamically simulate the impact range of each risk point. The boundary data of the impact range is obtained from the simulation results to determine the risk impact range list. Based on the risk impact range list, a local adjustment mechanism is used to replan the parts of the execution plan involving risk points. The spatial distribution optimization tool is used to adjust the relevant paths and task allocations to obtain the adjusted plan fragments. The step S8 further includes seamlessly integrating the adjusted plan fragment with other parts of the original execution plan. If any task allocation conflicts are found during the integration process, the conflicts are resolved using a priority sorting tool to obtain a complete integrated plan. Based on the integrated plan, verification tools are used to conduct a comprehensive test of the plan's feasibility. If the test results show that there are implementation obstacles in any link, alternative paths are extracted from the backup plan database to determine the final investigation plan; For the final survey plan, the plan content is correlated and matched with the actual environmental information of the field survey to obtain environmental adaptability feedback and obtain optimized execution guidance.

2. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: Said S1 comprises extracting an initial feature data set from geological information by constructing a three-dimensional spatial data processing framework, decomposing the original data by a multi-level grid partitioning method in view of the data distribution characteristics under a complex terrain environment, and obtaining a sub-region data set and a classification information set; decomposing the original data into a plurality of sub-region data sets by a multi-level grid partitioning method, and obtaining the classification results of local geological information in view of the data distribution characteristics within the sub-region set; optimizing and adjusting the classification information by using a support vector machine algorithm based on the classification results of the sub-region data sets in combination with the data distribution characteristics under a complex terrain environment, and determining a geological information set; if there is uneven data distribution in the geological information set In the case of the above, the sub-region set is supplemented by the data interpolation method to obtain a geological information data set with uniform distribution; based on the uniformly distributed geological information data set, the spatial interpolation technology is used to process the classification information for the multi-level network structure in three-dimensional space to obtain a geological information distribution map; through the geological information distribution map, combined with the initial feature data set, it is determined whether there are abnormal data points. If there are abnormal data points, they are smoothed to obtain the corrected geological information distribution result; based on the corrected geological information distribution result, the data fusion technology is used to integrate the classification information in the multi-level network according to the data characteristics under the terrain environment to determine the final geological information comprehensive data set.

3. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: The S2 includes using data processing tools to perform hierarchical analysis on the sub-region set based on the geological information and the results of the preliminary classification, obtaining the distribution characteristics of the outlier points and boundary information, and determining the preliminary anomaly distribution range; based on the anomaly distribution range, combined with in-depth analysis methods, local feature comparison is performed on the outlier points and boundary information to obtain an abnormal area distribution map; through the distribution map, a preset threshold is used to screen the abnormal area to obtain a set of candidate points for key features; for the candidate point set, combined with the distribution law of potential features, the support vector machine algorithm is used to classify and optimize the point set to determine the key geological feature points; if the key geological feature points are unevenly distributed in the sub-region set, the point set is adjusted by a data smoothing tool to obtain a feature point set with a balanced distribution; based on the evenly distributed feature point set, spatial mapping technology is used to associate and integrate the point set based on the spatial characteristics of the geological information to obtain a comprehensive geological feature distribution map; through the comprehensive geological feature distribution map, combined with the intermediate results of the screening process, the potential features are secondary verified to determine the final key geological feature point set.

4. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: The S3 includes, by means of data integration, constructing a feature association framework for the correspondence between the geological feature point set and the environmental parameters, and obtaining preliminary correspondence degree distribution data; based on the preliminary correspondence degree distribution data, screening is performed using a preset threshold, and if the correspondence degree is lower than the preset threshold, the point set is corrected to obtain a corrected point distribution data set; for the corrected point distribution data set, using a data comparison tool, analyzing the matching deviation between the point set and the environmental parameters, and determining a deviation point subset; based on the point subset, using a spatial adjustment method, positionally calibrating the point subset to obtain an adjusted point distribution structure; based on the adjusted point distribution structure, recalculating the correspondence degree between the point set and the environmental parameters through a feature association framework to obtain updated matching distribution data; based on the updated matching distribution data, if there are still points in the matching distribution data that do not reach the preset threshold, performing secondary optimization on these points through a data smoothing tool to obtain a final optimized distribution map; based on the final optimized distribution map, using spatial mapping technology to integrate the point set with the geological features to determine a comprehensive geological feature distribution view.

5. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: The S4 includes comparing the sampling points with the distribution map of geological features using a spatial analysis tool for the initial sampling point location set, determining the coverage ratio of the sampling points in the high-density area, and adjusting the sampling points if the coverage ratio does not reach a preset threshold to obtain an adjusted sampling point distribution data set; calculating the correlation between the sampling points and the feature distribution based on the adjusted sampling point distribution data set to analyze whether the point distribution in the high-density area is balanced; if the distribution deviation exceeds a preset range, performing local optimization on the points to obtain an optimized point distribution structure; processing the feature distribution in the grid division using a spatial interpolation method for the optimized point distribution structure, obtaining regional analysis data, and determining the boundary range of the density area. Based on the density zone boundary, the sampling points are rechecked using a data screening tool to determine whether all points fall within the boundary. If there are points that deviate from the boundary, the positions of the points that deviate from the boundary are adjusted to obtain a corrected sampling point set. For the corrected sampling point set, a data integration framework is used to correlate and map the location data with the distribution characteristics of geological features to obtain a comprehensive sampling point distribution view. Based on the comprehensive sampling point distribution view, a visualization tool is used to overlay and compare the sampling layout with the regional analysis results to determine the final sampling point configuration plan. For the final sampling point configuration plan, a data storage module is used to archive the location data and feature distribution information to obtain a structured sampling layout dataset.

6. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: S5 includes analyzing the matching degree between the path plan and the terrain using a data processing tool for the preliminary optimization of the navigation path plan. If the matching degree is lower than a preset threshold, the path plan is partially corrected to obtain a corrected path distribution. Based on the corrected path distribution, the correlation between the sampling points and the navigation path is calculated. If the correlation cannot cover the target area, the point distribution is redistributed to determine the adjusted point layout. For the adjusted point layout, an information integration framework is used to fuse the path plan with the travel cost data to obtain a comprehensive path cost distribution view. Based on the comprehensive path cost distribution view, a data screening tool is used to identify cost segments in the navigation path that are higher than the preset threshold. If the proportion of cost segments higher than the preset threshold exceeds the preset proportion range, alternative paths are planned for the cost segments higher than the preset threshold to obtain an optimized path structure. For the optimized path structure, a spatial interpolation method is used to analyze the coverage of the path plan and the sampling points to obtain regional distribution data. Based on the regional distribution data, the data verification module is used to evaluate the compliance of the path plan with the planning requirements. If the compliance does not meet the preset standard, the path structure is fine-tuned to determine the final navigation path configuration.

7. The geological mapping method based on UAV 3D modeling according to claim 1, characterized in that: The S7 includes using a pre-established resource allocation model for the data of the navigation path and the sampling location, analyzing the spatial distribution relationship between the sampling location and the path, and obtaining a preliminary resource allocation plan; based on the preliminary allocation plan, integrating the balance principle, using a spatial analysis tool to evaluate the distribution uniformity of the resource allocation, if the distribution uniformity is lower than a preset threshold, locally adjusting the resource allocation, and determining the adjusted configuration layout; based on the adjusted configuration layout, constructing a task allocation framework, decomposing the multi-objective task into multiple subtask units, and obtaining a set of decomposed task units; based on the set of decomposed task units, using a priority sorting method, calculating the matching degree between the subtask units and the resource allocation layout, and obtaining a task execution priority sequence; based on the task execution priority sequence, comparing the priority sequence with the overall requirements of the survey plan, if the comparison result shows that the task unit execution sequence does not meet the overall requirements, adjusting the priority sequence to determine the optimized execution sequence; based on the optimized execution sequence, using a data verification tool, analyzing the fit between the execution sequence and the complete execution plan, and obtaining the final field survey execution plan; For the final field survey execution plan, the execution plan is associated with the actual distribution of navigation paths and sampling locations through data mapping method to obtain the execution path distribution.

Images