Mine geological surveying and mapping system and method based on artificial intelligence

Abstract

The application relates to the technical field of surveying and mapping, and discloses a mine geological surveying and mapping system and method based on artificial intelligence, which comprises the following steps: substituting a sub-mine geological map of each to-be-surveyed mine sub-region into a twin neural network model to determine the geological type of each to-be-surveyed mine sub-region, comparing the geological type with a historical geological type, judging whether the geological type is changed, establishing a geological mark for the corresponding to-be-surveyed mine sub-region according to the geological condition of the target geological type, determining a radar scanning angle based on the sequence of the geological mark, determining whether to adjust the radar scanning angle based on a historical scanning set, when radar scanning is carried out on a changed geological mark or a multi-source geological mark, determining a compensation scheme of a radar reflection signal according to a signal-to-noise ratio, and completing the surveying and mapping of each to-be-surveyed mine sub-region based on the target radar scanning angle or the compensation scheme of the radar reflection signal, so that the stability and reliability of mine geological surveying and mapping are ensured.

Description

Technical Field

[0001] This invention relates to the field of surveying and mapping technology, and more specifically, to a mine geological surveying and mapping system and method based on artificial intelligence. Background Technology

[0002] Mine geological mapping is a crucial link in the rational development of mineral resources, ensuring safe production, and assessing the ecological environment. The accuracy and timeliness of mapping directly impact mine planning and operation. However, due to the complexity and variability of the mine geological environment, different areas may exhibit geological features such as rock strata fractures, variations in ore and rock properties, and differences in aquifer distribution, which places higher demands on mapping. In recent years, with the rapid development of UAV and radar detection technologies, the combined mapping mode has been widely used in the field of mine geological mapping due to its advantages such as operational flexibility, wide coverage, and less susceptibility to terrain limitations. However, UAV and radar detection are usually based on parameters set by human experience, which cannot be adjusted according to the complexity and variability of the mine geological environment. This leads to reduced mapping accuracy in geologically complex or changing areas, resulting in insufficient reliability of the mapping results.

[0003] Therefore, it is necessary to design an artificial intelligence-based mine geological mapping system and method to solve the problems existing in the current technology. Summary of the Invention

[0004] In view of this, the present invention proposes a mine geological mapping system and method based on artificial intelligence, which aims to solve the problem that the mapping of UAVs and radar detection is usually carried out by parameters set by human experience, which cannot be adjusted according to the complexity and variability of the mine geological environment. This leads to reduced mapping accuracy in areas with complex or changing geology, resulting in insufficient reliability of the mapping results.

[0005] In one aspect, this invention proposes an artificial intelligence-based method for mine geological mapping, comprising:

[0006] Obtain the geological map of the mine to be surveyed, divide the mine to be surveyed into several sub-regions based on the geological map, and substitute the sub-geological map of each sub-region into the twin neural network model to determine the geological type of each sub-region.

[0007] The geological type is compared with the historical geological type to determine whether the geological type has changed. When it is determined that there has been a change, the target geological type is determined based on the relationship between the adjacent mine sub-region to be surveyed and the mine sub-region to be surveyed. The target geological type is analyzed, and a geological marker is established for the corresponding mine sub-region to be surveyed based on the geological conditions of the target geological type.

[0008] The radar scanning angle is determined based on the order of the geological markers. When radar scanning is performed on an independent geological marker, it is determined whether to adjust the radar scanning angle based on the historical scan set. When it is determined that the radar scanning angle should be adjusted, the target radar scanning angle is determined based on the traversal result of the historical scan set. When radar scanning is performed on a changed geological marker or a multi-source geological marker, the signal-to-noise ratio of the radar reflection signal is determined based on the radar scanning angle, and a compensation scheme for the radar reflection signal is determined based on the signal-to-noise ratio.

[0009] The mapping of each mine sub-area to be mapped is completed based on the compensation scheme of the target radar scanning angle or the radar reflection signal.

[0010] Furthermore, when substituting the sub-mine geological map of each sub-region of the mine to be mapped into the twin neural network model to determine the geological type of each sub-region of the mine to be mapped, the following steps are taken:

[0011] Obtain a geological atlas of the mine and divide it into a training set and a test set. Use grid search to find the model parameters of the twin neural network model, establish the twin neural network model, fit the twin neural network model according to the training set, and substitute the test set into the twin neural network model to determine the prediction accuracy. When the prediction accuracy is greater than or equal to the prediction accuracy threshold, determine the geological type of each sub-region of the mine to be mapped according to the sub-geological map of each sub-region of the mine to be mapped.

[0012] Furthermore, when comparing the geological type with historical geological types to determine whether the geological type has changed, the process includes:

[0013] When the geological type is the same as the historical geological type, it is determined that the geological type has not changed, and the geological type is identified as the target geological type;

[0014] When the geological type is not the same as the historical geological type, it is determined that the geological type has changed.

[0015] Furthermore, when a change is determined, the determination of the target geological type based on the relationship between the adjacent sub-regions of the mine to be surveyed and the sub-regions of the mine to be surveyed includes:

[0016] Identify the sub-regions of mines to be surveyed where the geological type has changed, and obtain several adjacent sub-regions of mines to be surveyed in adjacent orientations;

[0017] Based on the association rule algorithm, several candidate item sets of the sub-region of mine to be surveyed that has undergone changes and its adjacent sub-regions to be surveyed are determined. Frequent itemsets are determined based on the support of each candidate item set. Based on the frequent itemsets, the association results between the sub-region of mine to be surveyed that has undergone changes and its adjacent sub-regions to be surveyed are determined.

[0018] Based on the correlation results, the target geological type of the sub-region of the mine to be surveyed that has undergone changes is determined.

[0019] Furthermore, when analyzing the target geological type and establishing geological markers for the corresponding sub-areas of the mine to be mapped based on the geological conditions of the target geological type, the process includes:

[0020] The geological identifiers include independent geological identifiers, altered geological identifiers, and multi-source geological identifiers;

[0021] When the target geological type is determined to be a geological type that does not change, and the target geological type is a single geological type, then the independent geological identifier is established for the corresponding sub-area of ​​the mine to be surveyed for the target geological type.

[0022] When the target geological type is determined to be one that does not change, and the target geological type is one of several types, then the multi-source geological identifier is established for the corresponding sub-area of ​​the mine to be surveyed for the target geological type.

[0023] When the target geological type is determined to be a change, and the target geological type is one of several types, the change geological identifier is established for the corresponding sub-area of ​​the mine to be surveyed for that target geological type.

[0024] Furthermore, when determining the radar scanning angle based on the order of the geological markers, and when performing radar scanning on independent geological markers, determining whether to adjust the radar scanning angle based on the historical scan set includes:

[0025] The radar scanning angles for the corresponding sub-regions of the mine to be mapped are determined according to the order of the independent geological markers.

[0026] The historical scan set includes historical radar qualified scan angles with historical independent geological identifiers, several historical radar scan angles, and several angle adjustment factors, with each historical radar scan angle corresponding to an angle adjustment factor.

[0027] If the radar scanning angle is greater than or equal to the historical qualified radar scanning angle, then it is determined that the radar scanning angle will not be adjusted.

[0028] If the radar scanning angle is less than the historical acceptable radar scanning angle, then it is determined that the radar scanning angle should be adjusted.

[0029] Furthermore, when determining to adjust the radar scanning angle, the process of determining the target radar scanning angle based on the traversal results of the historical scan set includes:

[0030] When there is a historical radar scanning angle in the historical scan set that is equal to the radar scanning angle, the radar scanning angle is adjusted by the angle adjustment factor corresponding to the historical radar scanning angle to determine the target radar scanning angle.

[0031] When there is no historical radar scanning angle in the historical scan set that is equal to the radar scanning angle, the radar scanning angle and the historical scan set are used as the dataset to be clustered. The angle adjustment factor corresponding to each historical radar scanning angle in the dataset to be clustered is extracted, the expected number of clusters k is determined to be 2, and the parameters of the Gaussian distribution are initialized to determine the responsibility value. Based on the responsibility value, a similar set corresponding to the radar scanning angle is obtained. The mean of the angle adjustment factor in the similar set is used to adjust the radar scanning angle to determine the target radar scanning angle.

[0032] The radar scanning angle and the angle adjustment factor are directly proportional.

[0033] Furthermore, when performing radar scanning with altered geological markers or multi-source geological markers, the method for determining the signal-to-noise ratio (SNR) of the radar reflected signal based on the radar scanning angle, and determining a compensation scheme for the radar reflected signal based on the SNR, includes:

[0034] The power of the radar reflected signal is determined, a noisy radar echo signal is collected and the total signal power is determined, the difference between the total signal power and the power of the radar reflected signal is obtained, and the difference is determined as the noise power;

[0035] The signal-to-noise ratio (SNR) of the radar reflected signal is determined based on the power and noise power of the radar reflected signal. The SNR is then compared with an SNR threshold, and a compensation scheme for the radar reflected signal is determined based on the comparison result.

[0036] Furthermore, when determining the compensation scheme for the radar reflection signal based on the comparison results, the following steps are included:

[0037] When the signal-to-noise ratio is less than or equal to the signal-to-noise ratio threshold, multi-pulse energy superposition is used to decompose the original single-pulse signal into several sub-pulses, and the radar signals are sent out in staggered time sequences.

[0038] When the signal-to-noise ratio is greater than the signal-to-noise ratio threshold, the beam divergence angle is reduced, and phase difference interference is performed on two consecutive radar reflection signals.

[0039] Compared with existing technologies, the beneficial effects of this invention are as follows: It intelligently determines the geological type of each sub-region of the mine to be surveyed through a twin neural network model, avoiding the subjectivity of relying on human experience. It uses artificial intelligence to learn and accurately extract and match geological maps of the mine, achieving standardized identification of geological types and laying a data foundation for surveying. The comparison between geological types and historical geological types enables real-time detection of geological changes in the sub-regions to be surveyed. By combining the relationship between adjacent sub-regions to be surveyed, the target geological type is determined and a dynamic geological marker is established, thereby dynamically tracking geological changes and avoiding surveying deviations caused by sudden geological changes. The radar scanning angle is dynamically adjusted for independent geological markers, and signal interference from changed or multi-source geological markers is handled through a signal-to-noise ratio compensation scheme, ensuring the adaptability of the scanning surveying parameters to complex geological conditions and improving the surveying accuracy of complex and variable areas, thus guaranteeing the stability and reliability of the surveying process.

[0040] On the other hand, this application also provides an artificial intelligence-based mine geological mapping system for applying the aforementioned artificial intelligence-based mine geological mapping method, including:

[0041] The data acquisition and analysis unit is configured to acquire a geological map of the mine to be surveyed, divide the mine to be surveyed into several sub-regions based on the geological map, and input the geological map of each sub-region into a twin neural network model to determine the geological type of each sub-region.

[0042] The first processing unit is configured to compare the geological type with the historical geological type, determine whether the geological type has changed, and when it is determined that there has been a change, determine the target geological type based on the relationship between the adjacent mine sub-region to be surveyed and the mine sub-region to be surveyed, analyze the target geological type, and establish a geological identifier for the corresponding mine sub-region to be surveyed based on the geological conditions of the target geological type.

[0043] The second processing unit is configured to determine the radar scanning angle based on the order of the geological markers; when performing radar scanning on an independent geological marker, determine whether to adjust the radar scanning angle based on the historical scan set; when it is determined that the radar scanning angle should be adjusted, determine the target radar scanning angle based on the traversal result of the historical scan set; when performing radar scanning on a changed geological marker or a multi-source geological marker, determine the signal-to-noise ratio of the radar reflection signal based on the radar scanning angle, and determine the compensation scheme for the radar reflection signal based on the signal-to-noise ratio.

[0044] The geological mapping unit is configured to complete the mapping of each mine sub-area to be mapped based on the target radar scanning angle or the radar reflection signal compensation scheme.

[0045] It is understandable that the aforementioned AI-based mine geological mapping system and method have the same beneficial effects, and will not be elaborated upon here. Attached Figure Description

[0046] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0047] Figure 1 A flowchart illustrating an artificial intelligence-based mine geological mapping method provided in an embodiment of the present invention;

[0048] Figure 2 This is a functional block diagram of an artificial intelligence-based mine geological mapping system provided in an embodiment of the present invention. Detailed Implementation

[0049] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey its scope to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0050] See Figure 1 As shown in some embodiments of this application, an artificial intelligence-based mine geological mapping method includes:

[0051] S100: Obtain the geological map of the mine to be surveyed, divide the mine to be surveyed into several sub-regions based on the geological map, and substitute the sub-geological map of each sub-region into the twin neural network model to determine the geological type of each sub-region.

[0052] S200: Compare geological types with historical geological types to determine whether there has been a change in geological type. If a change is determined, determine the target geological type based on the relationship between adjacent mine sub-regions to be surveyed and the target geological type. Analyze the target geological type and establish geological markers for the corresponding mine sub-regions to be surveyed based on the geological conditions of the target geological type.

[0053] S300: The radar scanning angle is determined based on the order of geological markers. When radar scanning is performed on independent geological markers, it is determined whether to adjust the radar scanning angle based on the historical scan set. When it is determined to adjust the radar scanning angle, the target radar scanning angle is determined based on the traversal results of the historical scan set. When radar scanning is performed on changed geological markers or multi-source geological markers, the signal-to-noise ratio of the radar reflection signal is determined based on the radar scanning angle, and a compensation scheme for the radar reflection signal is determined based on the signal-to-noise ratio.

[0054] S400: Based on the compensation scheme of the target radar scanning angle or radar reflection signal, it completes the mapping of each mine sub-area to be mapped.

[0055] Specifically, a geological map of the mine to be surveyed is obtained. This geological map is determined by 3D maps, GPS positioning, etc., and reflects the preliminary geological conditions of the mine. Based on the geological map, the mine is divided into several sub-regions. The number of sub-regions can be dynamically set according to the actual size of the mine; in this embodiment, 50 sub-regions are preferred. A Siamese neural network model is used to compare the geological maps of each sub-region, and the geological type of each sub-region is determined through matching within the Siamese neural network model. Siamese neural networks excel at capturing subtle feature differences and can accurately identify complex geological features such as rock strata fractures and aquifer distribution, providing a basic classification basis for subsequent mapping. By comparing the current geological type with historical data, it is generally believed that geology is affected by the surrounding area during migration or evolution. If changes occur, the geological correlation between adjacent sub-regions of the mine to be mapped is combined to determine the target geological type, and a geological label covering features such as rock strata structure and mineral composition is established for it, thereby enabling a quantitative assessment of geological changes in different regions. For independent geological label areas, the optimal angle is analyzed by traversing the historical scan set, thereby dynamically adjusting the radar scan angle. The historical scan set contains data from successful scans at different historical periods. For changed geological labels or multi-source geological labels, the signal-to-noise ratio of the radar reflection signal is calculated to determine the compensation scheme for the radar reflection signal to ensure signal quality.

[0056] Understandably, establishing different geological markers to address the impact of various geological changes allows for dynamic adjustment of radar scanning angles or determination of radar reflection signal compensation schemes. This enables the mapping process to effectively respond to different geological changes, ensuring the stability and reliability of the mapping. By using the determined target radar scanning angle or radar reflection signal compensation scheme, the mapping of each sub-area of ​​the mine to be mapped is completed, avoiding the uncertainty and subjectivity of human experience. This allows the mapping process to promptly capture changes such as rock strata movement and the emergence of new faults. For relatively complex changes in geological markers or multi-source geological markers, the signal-to-noise ratio compensation scheme can effectively eliminate signal distortion caused by abrupt changes in mineral and rock properties. This achieves closed-loop control of geological identification, change tracking, and parameter adjustment, while balancing mapping accuracy and efficiency, and improving the stability and reliability of the mapping results.

[0057] In some embodiments of this application, when substituting the sub-geological map of each mine sub-region to be mapped into a Siamese neural network model to determine the geological type of each mine sub-region to be mapped, the process includes: acquiring a mine geological atlas and dividing the mine geological atlas into a training set and a test set; using grid search to find the model parameters of the Siamese neural network model; establishing a Siamese neural network model; fitting the Siamese neural network model according to the training set; substituting the test set into the Siamese neural network model and determining the prediction accuracy; and when the prediction accuracy is greater than or equal to the prediction accuracy threshold, determining the geological type of each mine sub-region to be mapped based on the sub-geological map of each mine sub-region to be mapped.

[0058] Specifically, the mine geological atlas contains image data of all geological formations in the mine, along with corresponding geological types (such as mudstone, coal seams, granite, and quartzite). The atlas is divided into training and testing sets, typically in a 3:2 ratio, to ensure both sets contain diverse data, thereby improving the model's generalization ability. Grid search exhaustively searches for parameter combinations in the parameter space to find the model parameters for the Siamese Neural Network (SNN). A SNN is a neural network architecture used to compare the similarity or difference of inputs. It contains multiple layers of neurons of different types and activation functions, designed to capture complex relationships in the data. The training set is used to fit the SNN model, while the testing set is used to evaluate the fitted model, reducing the risk of overfitting and improving accuracy and stability. During the fitting process, the model attempts to learn patterns and relationships in the data to improve its prediction or classification ability. Prediction accuracy reflects the model's performance on unknown data and is a crucial indicator for evaluating model performance. The prediction accuracy threshold is preferably 0.8. By continuously fitting the parameters of the Siamese neural network model with the training set and the test set, the prediction accuracy of the model is made to be greater than or equal to the prediction accuracy threshold. This enables the accurate output of the geological type of each mine sub-region to be mapped, ensuring the stability and reliability of the mapping process and laying the foundation for subsequent judgment and analysis.

[0059] In some embodiments of this application, when comparing geological types with historical geological types to determine whether there has been a change in geological type, the following steps are taken: when the geological type is the same as the historical geological type, it is determined that there has been no change in geological type, and the geological type is identified as the target geological type; when the geological type is not the same as the historical geological type, it is determined that there has been a change in geological type.

[0060] Specifically, by comparing the current geological type with historical geological types one by one, it is possible to quickly and intuitively determine whether the geology of each sub-region of the mine to be mapped has migrated or evolved. If the geological type is the same as the historical geological type, the stratigraphic structure, composition, and physical properties are considered to be basically stable, and no change is determined. If they are different, it indicates that recent geological processes, such as fault displacement or rock weathering, have led to substantial changes in the geological properties, and the geological type is determined to have changed. In this case, a more in-depth analysis of the area is needed to improve the mapping accuracy. On the one hand, the comparison process can meet the monitoring needs of engineering mapping. On the other hand, it avoids the need for in-depth mining or re-inference of massive amounts of data and avoids the uncertainty of relying on manually set parameters. Especially in areas with frequent or complex geological activities, it effectively reduces misjudgments and omissions, thereby enhancing the reliability and stability of the mapping results, and thus providing data support for mine resource assessment, safe production, and environmental protection.

[0061] In some embodiments of this application, when a change is determined, determining the target geological type based on the relationship between adjacent sub-regions of mines to be surveyed and the sub-regions to be surveyed includes: identifying the sub-regions of mines to be surveyed whose geological type has changed, and obtaining several adjacent sub-regions of mines to be surveyed in adjacent orientations; determining several candidate item sets of the sub-regions of mines to be surveyed with the change and the adjacent sub-regions to be surveyed according to an association rule algorithm; determining frequent itemsets based on the support of each candidate item set; determining the association results between the sub-regions of mines to be surveyed with the change and the adjacent sub-regions to be surveyed based on the frequent itemsets; and determining the target geological type of the sub-regions of mines to be surveyed with the change based on the association results.

[0062] Specifically, mine geology often exhibits regional continuity or trend changes, and statistical dependencies between neighboring areas can be captured through association rules. For a sub-region of a mine to be surveyed where the geological type has changed, eight adjacent sub-regions in eight adjacent directions (east, west, south, north, and diagonal directions) are obtained. For example, if the sub-regions to be surveyed are divided into a nine-square grid, with the first row and first column as number 1 and arranged sequentially, the second row and first column as number 4, the third row and first column as number 7, and the sub-region with change is number 5, then the adjacent sub-regions to be surveyed are numbers 1 to 4 and numbers 6 to 9. If the sub-region with change is number 1, then the adjacent sub-regions to be surveyed are numbers 2, 4, and 5. Their respective geological types are then collected, and the sub-region with change is included in the survey. The geological types of a mining sub-region and its adjacent sub-regions to be surveyed are considered as a transaction. The items in the transaction are the geological types of each sub-region. Using the Apriori algorithm in association rule algorithms, candidate itemsets of length 1 are first generated for all transactions. Then, the support of each candidate itemset is calculated from the bottom up, which is the proportion of itemsets that appear in all transactions. Itemets with a support greater than 2 are selected as frequent itemsets. Frequent itemsets essentially reveal the combination patterns of the mining sub-region to be surveyed that has undergone changes and its adjacent mining sub-regions to be surveyed. For example, when the adjacent sub-regions are sandstone and mudstone, the mining sub-region to be surveyed that has undergone changes is very likely to be sandstone. The association results are generated based on frequent itemsets (e.g., if adjacent sub-regions of mines to be surveyed have geological types A and B, then the sub-region to be surveyed will be changed to geological type C). Furthermore, as the data is continuously updated, the frequent itemsets and association rules will also be updated synchronously, enabling the surveying to cope with newly emerging geological evolution patterns, thereby having a strong response capability to sudden or local changes and improving the reliability of the overall surveying results.

[0063] In some embodiments of this application, when analyzing the target geological type and establishing geological identifiers for the corresponding sub-regions of the mine to be surveyed based on the geological conditions of the target geological type, the following steps are taken: the geological identifiers include independent geological identifiers, changed geological identifiers, and multi-source geological identifiers. When the target geological type is determined to be a geological type without changes and the target geological type is one type, an independent geological identifier is established for the corresponding sub-regions of the mine to be surveyed for that target geological type. When the target geological type is determined to be a geological type without changes and the target geological type is several types, a multi-source geological identifier is established for the corresponding sub-regions of the mine to be surveyed for that target geological type. When the target geological type is determined to be a geological type with changes and the target geological type is several types, a changed geological identifier is established for the corresponding sub-regions of the mine to be surveyed for that target geological type.

[0064] Specifically, independent geological markers reflect that the geological conditions of the sub-region of the mine to be mapped are stable and have simple characteristics, containing only one type of geology, such as coal seams. Multi-source geological markers reflect that the sub-region of the mine to be mapped is a mixed distribution of multiple geologies, such as a mixture of coal seams and granite. Change geological markers reflect whether the geological type has changed or whether it is a mixed distribution of multiple geologies. Furthermore, the sub-regions of the mine to be mapped with change geological markers are often located in areas where faults, sedimentation, and erosion intersect. By establishing three marker systems, the traceability and stability of the mapping results are enhanced, the ability to dynamically capture data relationships and metadata between different markers is strengthened, the subjectivity and uncertainty of human experience are avoided, and the risk of mapping errors caused by geological complexity is reduced. While ensuring mapping efficiency, the adaptability to complex geological environments and the reliability of mapping results are improved.

[0065] In some embodiments of this application, when determining the radar scanning angle based on the order of geological markers, and when performing radar scanning on independent geological markers, determining whether to adjust the radar scanning angle based on the historical scan set includes: determining the radar scanning angle for the corresponding mine sub-area to be mapped according to the order of independent geological markers; the historical scan set includes historical qualified radar scanning angles of historical independent geological markers, several historical radar scanning angles, and several angle adjustment factors, and each historical radar scanning angle corresponds to an angle adjustment factor; when the radar scanning angle is greater than or equal to the historical qualified radar scanning angle, it is determined that the radar scanning angle will not be adjusted; when the radar scanning angle is less than the historical qualified radar scanning angle, it is determined that the radar scanning angle will be adjusted.

[0066] Specifically, the radar scanning angles are determined for the corresponding mine sub-areas to be surveyed according to the order of independent geological markers. Taking the nine-square grid as an example again, when the mine sub-areas to be surveyed with independent geological markers are numbers 1 and 5, the radar scanning angles are determined starting with number 1, and the mine sub-areas to be surveyed are surveyed sequentially. This avoids the decrease in surveying stability caused by drones flying across the mine sub-areas with independent geological markers. The radar scanning angles are determined based on the height and width of the mine sub-area to be surveyed and according to the angle model. The training process of the angle model is the same as that of the Siamese neural network model, and will not be repeated here. Because independent geological markers reflect the stable and singular geological conditions of the sub-region of the mine to be mapped, containing only one type of geology, they do not require detailed measurement of the region compared to multi-source geological markers and modified geological markers. The historical scan set serves as a reference, including historically verified qualified radar scan angles (i.e., the lowest angle threshold for obtaining effective signals), historical scan angle samples, and corresponding angle adjustment factors (reflecting the degree of influence of angle deviation on signal quality). If the radar scan angle is greater than or equal to the historical qualified radar scan angle, it indicates that the radar scan angle can perform large-area scanning of the region to ensure the integrity of the mapping process. If the radar scan angle is less than the historical qualified radar scan angle, it indicates that the radar scan angle may lead to insufficient reflection due to insufficient angle (e.g., low-angle scanning of steep rock layers easily produces shadow areas). In this case, it is determined to adjust the radar scan angle, ensuring both scanning accuracy and efficiency, to leverage the stability advantage of the region corresponding to the independent geological marker, reduce interference and errors from human judgment, and enhance adaptability to different mapping conditions.

[0067] In some embodiments of this application, when determining the target radar scanning angle based on the traversal results of the historical scan set, the following steps are taken: when there is a historical radar scanning angle in the historical scan set that is equal to the radar scanning angle, the radar scanning angle is adjusted by the angle adjustment factor corresponding to the historical radar scanning angle to determine the target radar scanning angle; when there is no historical radar scanning angle in the historical scan set that is equal to the radar scanning angle, the radar scanning angle and the historical scan set are used as a dataset to be clustered, and the angle adjustment factor corresponding to each historical radar scanning angle in the dataset to be clustered is extracted, the expected number of clusters k is determined to be 2, and the parameters of the Gaussian distribution are initialized to determine the responsibility value. Based on the responsibility value, a similar set corresponding to the radar scanning angle is obtained, and the mean of the angle adjustment factor in the similar set is used to adjust the radar scanning angle to determine the target radar scanning angle. The radar scanning angle and the angle adjustment factor are directly proportional.

[0068] Specifically, when historical data matching the current radar scan angle is found, this data can be directly used to determine the angle adjustment factor, ensuring the reliability and consistency of the adjustment. For cases where the current radar scan angle does not perfectly match historical data, cluster analysis is used to identify similar sets to dynamically determine the angle adjustment factor, adapting to changes in different geological conditions. Data-driven automated adjustment reduces reliance on human experience and intuition, lowering the uncertainty and operational risks associated with human judgment, thereby improving the automation level of the surveying process and the accuracy of the results. Comprehensive utilization of a large amount of historical data provides rich reference information for determining the angle adjustment factor. Accurate angle adjustment factors avoid excessive or insufficient angle adjustments, thus optimizing surveying resources and reducing energy and material waste. Assuming the radar scan angle is U and the angle adjustment factor is F, the determined target radar scan angle is U*F. By establishing a direct proportional relationship between the radar scan angle and the angle adjustment factor, precise control of the scan angle is achieved in complex geological environments such as mines, ensuring the stability of the surveying process.

[0069] In some embodiments of this application, when performing radar scanning with altered geological markers or multi-source geological markers, determining the signal-to-noise ratio (SNR) of the radar reflected signal based on the radar scanning angle, and determining a compensation scheme for the radar reflected signal based on the SNR, includes: determining the power of the radar reflected signal, acquiring a noisy radar echo signal and determining the total signal power, obtaining the difference between the total signal power and the power of the radar reflected signal, determining the difference as noise power, determining the SNR of the radar reflected signal based on the power of the radar reflected signal and the noise power, comparing the SNR with a SNR threshold, and determining a compensation scheme for the radar reflected signal based on the comparison result.

[0070] In some embodiments of this application, when determining the compensation scheme for the radar reflection signal based on the comparison results, the following steps are included: when the signal-to-noise ratio is less than or equal to the signal-to-noise ratio threshold, multi-pulse energy superposition is used to decompose the original single-pulse signal into several sub-pulses, and the radar signals are sent out in a staggered time sequence; when the signal-to-noise ratio is greater than the signal-to-noise ratio threshold, the beam divergence angle is reduced, and phase difference interference is performed on two consecutive radar reflection signals.

[0071] Specifically, multi-source geological markers reflect that the sub-region of the mine to be mapped is a mixed distribution of multiple geological types, such as a mixture of coal seams and granite. Changed geological markers, on the other hand, reflect whether the geological type has changed or is still a mixed distribution of multiple geological types. Furthermore, the sub-regions of the mine to be mapped with changed geological markers are often located in areas where faults, sedimentation, and erosion intersect. Adjusting the radar scanning angle is not applicable to areas reflected by multi-source and changed geological markers because these areas exhibit highly complex geological structures and irregular variations in physical properties (such as density, dielectric constant, and conductivity). Adjusting the radar scanning angle relies on the assumption of uniform or regular geological interfaces. By adjusting the radar scanning angle to reduce interference and thus enhance the reflected signal, the mixing of multiple geological types and the interweaving of faults, sedimentation, and erosion create numerous irregular interfaces and fracture zones, resulting in significant differences in the reflection, refraction, and scattering characteristics of different geological types to radar waves. Simply adjusting the radar scanning angle cannot adapt to the signal response of all geological types, and may even mask the effective signals of some geological types or enhance interference signals. Therefore, the power of the radar reflected signal is first determined, then the radar echo signal with noise is collected and the total signal power is determined. The difference between the total signal power and the power of the radar reflected signal is obtained, and the difference is determined as the noise power. Finally, the signal-to-noise ratio is determined to achieve a quantitative assessment of the signal purity.

[0072] It is understandable that the signal-to-noise ratio (SNR) threshold can be dynamically set according to the intensity of the radar reflection signal in the actual mine. In this embodiment, the SNR threshold is preferably 10dB. When the SNR is less than or equal to the SNR threshold, it indicates that the radar reflection signal is greatly affected by geological clutter, such as multipath reflection in fault fracture zones. In this case, multi-pulse energy superposition is used to decompose a single pulse into multiple staggered sub-pulses. Random noise is canceled by energy accumulation. Multi-pulse superposition accumulates effective signals through time-staggered peak accumulation, which can improve the signal strength in low SNR areas and thus suppress sudden noise in geology. When the SNR is greater than the SNR threshold, it indicates that the radar reflection signal quality is good but the resolution needs to be improved. For example, multi-lithological transition zones require higher resolution. In this case, the radar beam divergence angle is reduced to focus energy, and phase difference interference is performed on continuous echoes. While ensuring the radar scanning speed, the ability to identify fine geological interfaces is enhanced, thereby adapting to the fine mapping needs of various geologies and ensuring the stability and reliability of the mapping process.

[0073] In summary, the beneficial effects of this invention are as follows: By using a twin neural network model to intelligently determine the geological type of each sub-region of the mine to be mapped, the subjectivity of relying on human experience is avoided. Artificial intelligence is used to learn and accurately extract and match geological maps of the mine, achieving standardized identification of geological types and laying a data foundation for surveying. Comparison of geological types with historical geological types enables real-time detection of geological changes in the sub-regions to be mapped. The relationship between adjacent sub-regions to be mapped determines the target geological type and establishes dynamic geological markers, thereby dynamically tracking geological changes and avoiding surveying deviations caused by sudden geological changes. Dynamic adjustment of the radar scanning angle for independent geological markers and processing of signal interference from changed or multi-source geological markers through a signal-to-noise ratio compensation scheme ensure the adaptability of the scanning surveying parameters to complex geological conditions, improving the surveying accuracy of complex and variable areas, and thus guaranteeing the stability and reliability of the surveying process.

[0074] In another preferred embodiment based on the above embodiments, see [reference] Figure 2 As shown, this embodiment provides an artificial intelligence-based mine geological mapping system for applying the aforementioned artificial intelligence-based mine geological mapping method, including:

[0075] The data acquisition and analysis unit is configured to acquire the geological map of the mine to be surveyed, divide the mine to be surveyed into several sub-regions based on the geological map, and input the sub-geological map of each sub-region into the twin neural network model to determine the geological type of each sub-region.

[0076] The first processing unit is configured to compare the geological type with the historical geological type to determine whether there is a change in the geological type. When a change is determined, the target geological type is determined based on the relationship between the adjacent mine sub-regions to be surveyed and the mine sub-regions to be surveyed. The target geological type is analyzed, and a geological identifier is established for the corresponding mine sub-regions to be surveyed based on the geological conditions of the target geological type.

[0077] The second processing unit is configured to determine the radar scanning angle based on the order of geological markers. When performing radar scanning on an independent geological marker, it determines whether to adjust the radar scanning angle based on the historical scan set. When it is determined to adjust the radar scanning angle, it determines the target radar scanning angle based on the traversal results of the historical scan set. When performing radar scanning on a changed geological marker or a multi-source geological marker, it determines the signal-to-noise ratio of the radar reflection signal based on the radar scanning angle and determines the compensation scheme for the radar reflection signal based on the signal-to-noise ratio.

[0078] The geological mapping unit is configured to complete the mapping of each mine sub-area to be mapped based on a compensation scheme based on the target radar scanning angle or radar reflection signal.

[0079] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0080] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A mine geological mapping method based on artificial intelligence, characterized in that, The method comprises the following steps: obtaining a mine geological map of a mine to be surveyed, dividing the mine to be surveyed according to the mine geological map to determine a plurality of mine sub-regions to be surveyed, and determining the geological type of each mine sub-region to be surveyed by substituting the mine sub-region geological map of each mine sub-region to be surveyed into a twin neural network model; comparing the geological type with a historical geological type to determine whether the geological type has changed, determining a target geological type based on the relationship between adjacent mine sub-regions to be surveyed and the mine sub-region to be surveyed when it is determined that there is a change, analyzing the target geological type, and establishing a geological mark for the corresponding mine sub-region to be surveyed according to the geological conditions of the target geological type; determining a radar scanning angle based on the order of the geological mark, determining whether to adjust the radar scanning angle based on a historical scanning set when radar scanning is performed on an independent geological mark, determining a target radar scanning angle based on the traversal result of the historical scanning set when it is determined that the radar scanning angle is adjusted, determining the signal-to-noise ratio of a radar reflection signal based on the radar scanning angle when radar scanning is performed on a change geological mark or a multi-source geological mark, and determining a compensation scheme for the radar reflection signal according to the signal-to-noise ratio; completing the surveying of each mine sub-region to be surveyed based on the target radar scanning angle or the compensation scheme for the radar reflection signal; when analyzing the target geological type and establishing a geological mark for the corresponding mine sub-region to be surveyed according to the geological conditions of the target geological type, the method comprises the following steps: the geological mark comprises the independent geological mark, the change geological mark, and the multi-source geological mark; when the target geological type is the geological type without change, and the target geological type is one geological type, the independent geological mark is established for the corresponding mine sub-region to be surveyed of the target geological type; when the target geological type is the geological type without change, and the target geological type is a plurality of geological types, the multi-source geological mark is established for the corresponding mine sub-region to be surveyed of the target geological type; when the target geological type is the geological type with change, and the target geological type is a plurality of geological types, the change geological mark is established for the corresponding mine sub-region to be surveyed of the target geological type.

2. The mine geological mapping method based on artificial intelligence according to claim 1, characterized in that, when the geological type of each mine sub-region to be surveyed is determined by substituting the mine sub-region geological map of each mine sub-region to be surveyed into a twin neural network model, the method comprises the following steps: obtaining a mine geological map set, dividing the mine geological map set into a training set and a test set, finding the model parameters of the twin neural network model using grid search, establishing the twin neural network model, fitting the twin neural network model according to the training set, substituting the test set into the twin neural network model to determine the prediction accuracy, and determining the geological type of each mine sub-region to be surveyed according to the mine sub-region geological map of each mine sub-region to be surveyed when the prediction accuracy is greater than or equal to a prediction accuracy threshold. 3.The mine geological mapping method based on artificial intelligence according to claim 2, characterized in that, when the geological type is compared with the historical geological type to determine whether the geological type has changed, the method comprises the following steps: When the geological type is the same as the historical geological type, it is determined that the geological type is not changed, and the geological type is determined as a target geological type; When the geological type is not the same as the historical geological type, it is determined that the geological type is changed. 4.The mine geological mapping method based on artificial intelligence according to claim 3, characterized in that, When it is determined that there is a change, the target geological type is determined based on the relationship between the adjacent to-be-mapped mine sub-regions and the to-be-mapped mine sub-region, comprising: determining the to-be-mapped mine sub-region in which the geological type is changed, and obtaining a plurality of adjacent to-be-mapped mine sub-regions in adjacent directions; determining a plurality of candidate item sets of the to-be-mapped mine sub-region in which the geological type is changed and the adjacent to-be-mapped mine sub-regions according to an association rule algorithm, determining a frequent item set according to the support degree of each candidate item set, and determining an association result of the to-be-mapped mine sub-region in which the geological type is changed and the adjacent to-be-mapped mine sub-regions based on the frequent item set; determining the target geological type of the to-be-mapped mine sub-region in which the geological type is changed based on the association result. 5.The mine geological mapping method based on artificial intelligence according to claim 4, characterized in that, When the radar scanning angle is determined based on the order of the geological markers, and when it is determined whether to adjust the radar scanning angle based on a historical scanning set when radar scanning is performed on an independent geological marker, comprising: determining the radar scanning angle for the corresponding to-be-mapped mine sub-region according to the order of the independent geological markers; the historical scanning set comprises a historical radar qualified scanning angle of a historical independent geological marker, a plurality of historical radar scanning angles, and a plurality of angle adjustment factors, and each historical radar scanning angle corresponds to an angle adjustment factor; when the radar scanning angle is greater than or equal to the historical radar qualified scanning angle, it is determined that the radar scanning angle is not adjusted; when the radar scanning angle is less than the historical radar qualified scanning angle, it is determined that the radar scanning angle is adjusted. 6.The mine geological mapping method based on artificial intelligence according to claim 5, characterized in that, When it is determined that the radar scanning angle is adjusted, the target radar scanning angle is determined based on the traversal result of the historical scanning set, comprising: when there is a historical radar scanning angle equal to the radar scanning angle in the historical scanning set, the angle adjustment factor corresponding to the historical radar scanning angle is used to adjust the radar scanning angle to determine the target radar scanning angle; when there is no historical radar scanning angle equal to the radar scanning angle in the historical scanning set, the radar scanning angle and the historical scanning set are used as a to-be-clustered data set, the angle adjustment factor corresponding to each historical radar scanning angle in the to-be-clustered data set is extracted, the expected cluster number k is determined as 2, the responsibility value of the parameter of the Gaussian distribution is initialized, the similar set corresponding to the radar scanning angle is obtained according to the responsibility value, the mean value of the angle adjustment factor in the similar set is used to adjust the radar scanning angle to determine the target radar scanning angle; the radar scanning angle and the angle adjustment factor are in a proportional relationship. 7.The mine geological mapping method based on artificial intelligence according to claim 6, characterized in that, When radar scanning is performed on a changed geological marker or a multi-source geological marker, the signal-to-noise ratio of the radar reflection signal is determined based on the radar scanning angle, and the compensation scheme of the radar reflection signal is determined according to the signal-to-noise ratio, comprising: Determine the power of the radar reflection signal, collect the radar echo signal with noise and determine the total power of the signal, obtain the difference between the total power of the signal and the power of the radar reflection signal, and determine the difference as the noise power; Determine the signal-to-noise ratio of the radar reflection signal based on the power of the radar reflection signal and the noise power, compare the signal-to-noise ratio with the signal-to-noise ratio threshold, and determine the compensation scheme of the radar reflection signal according to the comparison result. 8.The mine geological mapping method based on artificial intelligence according to claim 7, characterized in that, When determining the compensation scheme of the radar reflection signal according to the comparison result, it includes: When the signal-to-noise ratio is less than or equal to the signal-to-noise ratio threshold, the multi-pulse energy superposition is adopted, the original single-pulse signal is decomposed into several sub-pulses, and the radar signal is sent in time sequence. When the signal-to-noise ratio is greater than the signal-to-noise ratio threshold, the beam divergence angle is reduced, and the phase difference interference is performed on the continuous two radar reflection signals.

9. A mine geological mapping system based on artificial intelligence, for applying the mine geological mapping method based on artificial intelligence according to any one of claims 1-8, characterized in that, It includes: The acquisition and analysis unit is configured to obtain the mine geological map of the mine to be surveyed, divide the mine to be surveyed according to the mine geological map to determine a plurality of mine sub-regions to be surveyed, and determine the geological type of each mine sub-region to be surveyed by substituting the mine sub-region geological map into the twin neural network model. The first processing unit is configured to compare the geological type with the historical geological type, judge whether the geological type is changed, determine the target geological type based on the relationship between the adjacent mine sub-region to be surveyed and the mine sub-region to be surveyed when it is determined that the geological type is changed, analyze the target geological type, and establish the geological identification of the corresponding mine sub-region to be surveyed according to the geological condition of the target geological type. The second processing unit is configured to determine the radar scanning angle based on the order of the geological identification, determine whether to adjust the radar scanning angle based on the historical scanning set when the radar scanning is performed in the independent geological identification, determine the target radar scanning angle based on the traversal result of the historical scanning set when it is determined to adjust the radar scanning angle, determine the signal-to-noise ratio of the radar reflection signal based on the radar scanning angle when the radar scanning is performed in the changed geological identification or the multi-source geological identification, and determine the compensation scheme of the radar reflection signal according to the signal-to-noise ratio. The geological surveying unit is configured to complete the surveying of each mine sub-region to be surveyed based on the target radar scanning angle or the compensation scheme of the radar reflection signal.

Images