A geophysical anomaly data management system and method for gold exploration

Abstract

This invention discloses a geophysical anomaly data management system and method for gold mine exploration, relating to the field of data analysis technology. The system includes: a historical data management module, a feature extraction and analysis module, a validity assessment model construction module, and a real-time data acquisition and validity assessment module. The historical data management module divides historical geophysical anomaly data into historical valid data and historical invalid data. The feature extraction and analysis module analyzes the deviation between historical valid data and historical invalid data, extracts and marks key validity features. The validity assessment model construction module constructs a comprehensive validity assessment model based on the key validity features and calculates a comprehensive validity assessment index threshold. The real-time data acquisition and validity assessment module acquires real-time geophysical anomaly data, extracts key validity features, calculates a real-time comprehensive validity assessment index, and identifies real-time valid data.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a geophysical anomaly data management system and method for gold mine exploration. Background Technology

[0002] Gold exploration plays a crucial role in mineral resource exploration, particularly in the research of gold mineralization and prospecting models, where significant progress has been made. However, with increasing exploration depth, the detection depth of traditional surface geophysical and geochemical methods is gradually limited, revealing their limitations in the exploration of deep or concealed ore bodies. The effectiveness of surface anomaly signals in deep ore body detection has significantly decreased, and traditional geophysical and geochemical methods struggle to provide sufficient information to support this process. The detection and exploration of deep and concealed ore bodies typically require more refined and in-depth geological, geophysical, and geochemical information. In this regard, deep boreholes provide more direct and effective prospecting information for gold exploration. Especially when probing deep ore bodies, borehole measuring points are closer to the target body, obtaining stronger response signals and possessing stronger anti-interference capabilities. Therefore, the application of deep borehole geophysical and geochemical technology is increasingly becoming a research hotspot in the field of gold exploration, especially in deep exploration, where the advantages of borehole geophysical and geochemical technology are becoming increasingly prominent.

[0003] However, geophysical anomaly data obtained through deep borehole geophysical exploration is affected by various factors, requiring exploration personnel to rely heavily on experience and repeated verification to accurately define anomaly areas. Currently, most gold exploration systems still rely on traditional data analysis methods, which often lack intelligent and automated analysis tools. Therefore, the analysis of the validity of geophysical anomaly data urgently needs the introduction of advanced intelligent analysis methods to improve analysis efficiency and accuracy. Summary of the Invention

[0004] The purpose of this invention is to provide a geophysical anomaly data management system and method for gold mine exploration, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A method for managing geomorphic anomaly data in gold exploration includes the following steps:

[0007] Step S100. Obtain historical geophysical anomaly data and historical event records from the database within the area where the gold mine exploration task has been completed. Based on the historical event records, divide the historical geophysical anomaly data into historical valid data and historical invalid data.

[0008] Step S200. By analyzing the deviation relationship between historical valid data and historical invalid data, the distinguishing features between the two are identified, and the distinguishing features between historical valid data and historical invalid data are marked as key validity features;

[0009] Step S300. Based on the key features of effectiveness, and combining historical effective data and historical invalid data, construct a comprehensive evaluation model for the effectiveness of geomorphic anomaly data, thereby obtaining a comprehensive effectiveness evaluation index; calculate the corresponding comprehensive effectiveness evaluation index based on historical effective data and historical invalid data, thereby obtaining a comprehensive effectiveness evaluation threshold.

[0010] Step S400. Collect real-time geophysical anomaly data of the gold exploration area and extract key features of effectiveness from the real-time geophysical anomaly data; calculate the real-time effectiveness comprehensive evaluation index based on the key features of effectiveness and in combination with the effectiveness comprehensive evaluation model; compare the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold, and extract real-time effective data from the real-time geophysical anomaly data based on the comparison results.

[0011] Furthermore, step S100 includes:

[0012] The historical event records refer to the records of historical geophysical anomalies within areas where gold exploration tasks have been completed. These records are used to determine the validity of historical geophysical anomaly data by comparing it with actual usage. Based on the historical event records, the historical geophysical anomaly data is divided into historically valid data and historically invalid data, and historically valid dataset A and historically invalid dataset B are constructed. Historically valid dataset A and historically invalid dataset B are represented as follows: A = {a1, a2, ..., an}, B = {b1, b2, ..., bm}, where A represents the historically valid dataset, a1 represents the first element of historically valid dataset A, a2 represents the second element of historically valid dataset A, and so on, with an representing the nth element of historically valid dataset A; B represents the historically invalid dataset, b1 represents the first element of historically invalid dataset B, b2 represents the second element of historically invalid dataset B, and so on, with bm representing the mth element of historically invalid dataset B.

[0013] Historical geophysical anomaly data refers to data detected during past geophysical, geochemical, and geological exploration processes that show significant differences from the normal background values ​​at the time. Specifically, geophysical anomaly data includes geological exploration anomaly data, geophysical anomaly data, and geochemical anomaly data. Geological exploration anomaly data refers to data discovered during geological exploration that is significantly different from the surrounding geological environment or expected geological features. This may include sudden changes in rock type, anomalies in stratigraphic structure, and abrupt changes in geological structures. These data often reveal the complexity of underground geological structures or the existence of special geological events. Geophysical anomaly data refers to data obtained during geophysical exploration by measuring geophysical fields (such as gravitational, magnetic, and electric fields). Geochemical anomaly data refers to data obtained through analysis of the content of elements or compounds in surface or underground samples during geochemical exploration. These data differ significantly from normal background values ​​and may manifest as gravity anomalies, magnetic anomalies, electrical anomalies, or abnormal changes in parameters such as seismic wave velocity and attenuation. They are usually related to the physical properties of underground geological bodies (such as density, magnetism, and electrical conductivity) and provide important information for inferring underground geological structures. Geochemical anomaly data refers to data that differs significantly from normal background values ​​and is found during geochemical exploration by analyzing the content of elements or compounds in surface or underground samples. These data may indicate the enrichment or depletion of certain elements or compounds and are usually related to the existence of underground ore bodies, geological tectonic activity, or migration of geochemical elements. They are important clues for mineral exploration and geological research.

[0014] Furthermore, step S200 includes:

[0015] S201. Based on historical valid dataset A and historical invalid dataset B, extract features from each element in historical valid dataset A and historical invalid dataset B respectively, and perform normalization processing, so that each element in historical valid dataset A and historical invalid dataset B is represented as a feature vector, and is represented as follows:

[0016] V_ai=[vai_1,vai_2,...,vai_p],V_bi=[vbi_1,vbi_2,...,vbi_p];

[0017] Where ai represents the i-th element in the historical valid dataset A, where i ranges from 1 to n; V_ai represents the feature vector corresponding to the i-th element in the historical valid dataset A, vai_1 represents the first-dimensional feature value of the feature vector V_ai, vai_2 represents the second-dimensional feature value of the feature vector V_ai, and so on, with vai_p representing the p-th dimension feature value of the feature vector V_ai, where p is the feature dimension, covering multiple indicators such as geology, geophysics, and geochemistry; similarly, V_bi represents the feature vector corresponding to the i-th element in the historical invalid dataset B, vbi_1 represents the first-dimensional feature value of the feature vector V_bi, vbi_2 represents the second-dimensional feature value of the feature vector V_bi, and so on, with vbi_p representing the p-th dimension feature value of the feature vector V_ai; due to the significant characteristics of geomorphic anomaly data such as high dimensionality and spatial distribution, commonly used algorithms include PCA, ICA, wavelet transform, autoencoder, SVM, K-means clustering, and neural networks. The specific algorithm chosen depends on the characteristics of the data (such as noise, dimensionality, spatial distribution, nonlinearity, etc.) and the requirements of the specific task. For example, if the data is noisy and requires dimensionality reduction, PCA or ICA may be more suitable; if it is necessary to capture multi-scale anomalies, wavelet transform may be more effective; for complex nonlinear relationships, neural networks and autoencoders may be better choices; the specific algorithm should be selected by the relevant personnel.

[0018] S202. For each element in the historical valid dataset A, calculate the difference between the feature values ​​of each dimension of the feature vectors V_ax and V_ay corresponding to two different elements ax and ay, thereby obtaining the minimum and maximum absolute values ​​of the difference between the feature values ​​of each dimension of the feature vectors in the historical valid dataset A, denoted as minΔvai_t and maxΔvai_t, respectively, where i takes values ​​from 1 to n and t takes values ​​from 1 to p; calculate the difference between the feature value of each dimension of the feature vector V_bi corresponding to each element in the historical invalid dataset B and the feature value of the corresponding dimension of V_ai corresponding to each element in the historical valid dataset A, thereby obtaining the absolute value of the difference Δvi_t; summarize the absolute values ​​of the difference Δvi_t of each dimension of the feature vector V_bi corresponding to each element in the historical invalid dataset B, thus forming the difference. Let Ct be a set, and Ct = {Δv1_t, Δv2_t, ..., Δvn_t}, where Δv1_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_a1 corresponding to the first element in the historical valid dataset A; Δv2_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_a2 corresponding to the second element in the historical valid dataset A; and so on, Δvn_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_an corresponding to the n-th element in the historical valid dataset A.

[0019] S203. For each element in the historical invalid dataset B, corresponding to the difference set Ct, compare the minimum and maximum absolute values ​​of the differences between each element in the difference set Ct and the feature values ​​of the corresponding dimension of the feature vector in the historical valid dataset A. Count the number N of elements in the difference set Ct that do not belong to the interval [minΔvai_t, maxΔvai_t]. If N / n ≥ d, then use the feature corresponding to the dimension of the difference set Ct as the distinguishing feature between the historical valid data and the historical invalid data, where d represents the proportion threshold. For each element in the historical invalid dataset B, corresponding to the difference set Ct, summarize the corresponding distinguishing features to form a distinguishing feature set Qt. Perform an intersection calculation on all distinguishing feature sets Qt, and mark the distinguishing features corresponding to the intersection calculation result as key validity features.

[0020] Furthermore, step S300 includes:

[0021] S301. For the key features of validity, search in the feature vectors corresponding to the historical valid data and historical invalid data to find the feature values ​​of the key features of validity in the feature vectors corresponding to the historical valid data and historical invalid data respectively, and calculate the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity, denoted as sk and mk respectively, where sk represents the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity k in the historical valid dataset A, and mk represents the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity k in the historical invalid dataset B; calculate the weight coefficient wk of the key features of validity, and the specific calculation formula is: wk=|sk-mk| / σk, where σk represents the standard deviation of the feature values ​​corresponding to the key features of validity; traverse all key features of validity to obtain the weight coefficients corresponding to all key features of validity;

[0022] S302. Based on the weight coefficients corresponding to each key validity feature and combined with the feature values ​​corresponding to the key validity features, construct a comprehensive evaluation model for the validity of geomorphic anomaly data, and the corresponding calculation formula is: E=∑ R k=1 (wk×vk); where E represents the comprehensive effectiveness evaluation index, R represents the number of key effectiveness features, and vk represents the feature value corresponding to the k-th key effectiveness feature, and the feature value corresponding to the key effectiveness feature is found from the historical effective dataset A or the historical invalid dataset B; according to the comprehensive effectiveness evaluation model, the comprehensive effectiveness evaluation index E corresponding to each element in the historical effective dataset A and the historical invalid dataset B is calculated, and according to the division of the historical effective dataset A and the historical invalid dataset B, the comprehensive effectiveness evaluation index set EA of the historical effective dataset A and the comprehensive effectiveness evaluation index set EB of the historical invalid dataset B are obtained respectively. The comprehensive effectiveness evaluation index threshold E0 is obtained based on the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB. The comprehensive effectiveness evaluation index threshold E0 is used to determine whether the comprehensive effectiveness evaluation model needs to be adjusted, and the corresponding adjustments are made according to the judgment result.

[0023] Furthermore, based on the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB, the comprehensive effectiveness evaluation index threshold E0 is obtained, and the specific analysis is as follows:

[0024] Represent the values ​​corresponding to the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB on a number line. Calculate the intersection of the comprehensive effectiveness evaluation index sets EA and EB. If the intersection result is an empty set, extract the minimum value EA_min and the maximum value EA_max from the comprehensive effectiveness evaluation index set EA, and the minimum value EB_min and the maximum value EB_max from the comprehensive effectiveness evaluation index set EB. If EA_min > EB_max, then EA_min is used as the comprehensive effectiveness evaluation index threshold E0; if EA_max < EB_min, then EA_max is used as the comprehensive effectiveness evaluation index threshold E0.

[0025] If the intersection calculation result is not an empty set, compare the maximum value EA_max in the comprehensive effectiveness evaluation index set EA with the maximum value EB_max in the comprehensive effectiveness evaluation index set EB. If EA_max ≤ EB_max, then the elements in the intersection region are successively used as the undetermined comprehensive effectiveness evaluation index threshold E1. Under the undetermined comprehensive effectiveness evaluation index threshold E1, the values ​​corresponding to true positive TP, false positive FP, true negative TN, and false negative FN are obtained in sequence. Among them, true positive TP represents the number of samples that were successfully and correctly classified as comprehensive effectiveness evaluation index within the intersection region; false positive FP represents the number of samples that were misclassified as comprehensive effectiveness evaluation index within the intersection region; and true negative TN represents the number of samples that were successfully and correctly classified as comprehensive ineffectiveness evaluation index outside the intersection region. False negatives (FN) represent the number of samples misjudged as invalidity comprehensive assessment index outside the intersection region; calculate the true positive rate (TPR) and false positive rate (FPR), where the specific calculation formulas are: TPR = TP / (TP + FN), FPR = FP / (FP + TN); summarize the true positive rate (TPR) and false positive rate (FPR) under each undetermined validity comprehensive assessment index threshold E1, and plot the ROC curve, with the horizontal axis of the ROC curve being FPR and the vertical axis being TPR; select the undetermined validity comprehensive assessment index threshold E1 corresponding to the point in the ROC curve that is farthest from the (0,1) point as the validity comprehensive assessment index threshold E0; similarly, if EB_max ≤ EA_max, analyze according to the analysis process for the case of EA_max ≤ EB_max, thereby obtaining the validity comprehensive assessment index threshold E0;

[0026] The effectiveness comprehensive evaluation model is determined based on the effectiveness comprehensive evaluation index threshold E0 to determine whether it needs adjustment, and corresponding adjustments are made based on the determination results, as detailed below:

[0027] If the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is an empty set, then the current comprehensive effectiveness evaluation model does not need to be adjusted. If the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not an empty set, and the intersection calculation results are all proper subsets of the comprehensive effectiveness evaluation index sets EA and EB, then the current comprehensive effectiveness evaluation model needs to be adjusted. The specific adjustment process is as follows:

[0028] An adaptive adjustment mechanism is used to dynamically adjust the weight coefficients of key effectiveness features. The specific calculation formula is: w'k = αk × wk + βk, where αk and βk represent the adjustment factor and bias term corresponding to key effectiveness feature k, respectively. The weight coefficients of all key effectiveness features are iterated to obtain several adjusted weight coefficients, which are then normalized to ensure that the sum of the weight coefficients of all adjusted key effectiveness features equals 1. The expression for the adjusted comprehensive effectiveness evaluation model is: E_end = ∑ R k=1 (w'k×vk); Based on the adjusted comprehensive effectiveness evaluation model, calculate the comprehensive effectiveness evaluation index E_end for each element in the historical valid dataset A and the historical invalid dataset B, thereby obtaining the corresponding comprehensive effectiveness evaluation index set EA_end and comprehensive effectiveness evaluation index set EB_end. The intervals represented by the adjusted comprehensive effectiveness evaluation index set EA_end and comprehensive effectiveness evaluation index set EB_end on the number axis do not intersect, thus obtaining the corresponding comprehensive effectiveness evaluation index threshold E0.

[0029] Furthermore, the specific analysis process for comparing the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold in step S400, and extracting real-time effective data from the real-time geophysical anomaly data based on the comparison results, is as follows:

[0030] When the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is empty, and EA_min > EB_max, then EA_min is used as the comprehensive effectiveness evaluation index threshold E0; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0, and when E' ≥ E0, the real-time materialized abnormal data is considered real-time valid data; when the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is empty, and EA_max < EB_min, then EA_max is used as the comprehensive effectiveness evaluation index threshold E0; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0, and when E' ≤ E0, the real-time materialized abnormal data is considered real-time valid data.

[0031] When the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not empty, and EA_max≤EB_max; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0. When E'≤E0, the real-time materialized abnormal data is considered real-time valid data; when the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not empty, and EB_max≤EA_max, the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0. When E'≥E0, the real-time materialized abnormal data is considered real-time valid data.

[0032] A geophysical anomaly data management system for gold exploration includes: a historical data management module, a feature extraction and analysis module, an effectiveness evaluation model construction module, and a real-time data acquisition and effectiveness evaluation module;

[0033] The historical data management module retrieves historical geomorphic anomaly data and historical event records from the database for areas where gold mine exploration tasks have been completed. Based on the historical event records, the historical geomorphic anomaly data is divided into historical valid data and historical invalid data.

[0034] The feature extraction and analysis module analyzes the deviation relationship between historical valid data and historical invalid data to identify the distinguishing features between the two, and marks the distinguishing features between historical valid data and historical invalid data as key validity features;

[0035] The effectiveness assessment model construction module constructs a comprehensive effectiveness assessment model for geomorphic anomaly data based on key effectiveness features and combining historical effective and invalid data, thereby obtaining a comprehensive effectiveness assessment index. Based on historical effective and invalid data, the corresponding comprehensive effectiveness assessment index is calculated, thereby obtaining the comprehensive effectiveness assessment threshold.

[0036] The real-time data acquisition and effectiveness assessment module collects real-time geophysical anomaly data of the gold mine exploration area and extracts key effectiveness features from the real-time geophysical anomaly data. Based on the key effectiveness features and combined with the effectiveness comprehensive assessment model, it calculates the real-time effectiveness comprehensive assessment index. The real-time effectiveness comprehensive assessment index is compared with the effectiveness comprehensive assessment threshold, and real-time effective data is extracted from the real-time geophysical anomaly data based on the comparison results.

[0037] The historical data management module includes a historical data acquisition unit and a historical data division and classification unit;

[0038] The historical data acquisition unit retrieves historical geomorphic anomaly data and corresponding historical event records from the database within the area where gold mine exploration tasks have been completed. The historical data division and classification unit judges the validity of geomorphic anomaly data based on historical event records, divides historical geomorphic anomaly data into valid data and invalid data, and constructs historical valid datasets and historical invalid datasets.

[0039] The feature extraction and analysis module includes a feature extraction and normalization unit and a feature difference analysis unit;

[0040] The feature extraction and normalization unit extracts features for each element in the historical valid dataset and the historical invalid dataset, and performs normalization processing to convert them into feature vectors; the feature difference analysis unit analyzes the feature differences between the historical valid dataset and the invalid dataset, thereby identifying the distinguishing features between the two and marking them as key features of validity.

[0041] The effectiveness evaluation model construction module includes a feature weight calculation unit and a comprehensive evaluation model construction unit;

[0042] The feature weight calculation unit calculates the weight coefficient of each key validity feature based on the differences between historical valid and invalid data; the comprehensive evaluation model construction unit uses the weight coefficients of each key validity feature to construct a comprehensive evaluation model for the validity of geomorphic anomaly data; the comprehensive validity evaluation index corresponding to historical geomorphic anomaly data is calculated through the comprehensive validity evaluation model, and the comprehensive validity evaluation threshold is determined.

[0043] The real-time data acquisition and effectiveness evaluation module includes a real-time data acquisition unit and a real-time effectiveness evaluation unit;

[0044] The real-time data acquisition unit collects real-time geophysical anomaly data of the gold mine exploration area and extracts key features of the real-time geophysical anomaly data to determine its validity. The real-time validity assessment unit calculates a comprehensive real-time validity assessment index based on the key features of the real-time data and compares it with a comprehensive validity assessment threshold to identify valid real-time data.

[0045] Compared with existing technologies, the beneficial effects of this invention are as follows: By analyzing the deviation between historical valid and invalid data, this invention can accurately distinguish the key features between the two and construct a comprehensive validity assessment model using these key validity features. This avoids the problems of relying on experience-based judgment and repeated verification in traditional methods, thus improving the accuracy of data judgment. Through feature extraction and normalization, each abnormal data point is represented as a feature vector, and the difference is calculated to identify the distinguishing features between abnormal and normal data. These distinguishing features are then used to construct a comprehensive validity assessment model. This multi-dimensional feature analysis, combined with weighted key validity features, can more accurately judge the validity of abnormal data, unlike traditional methods that rely solely on experience. By calculating and analyzing the intersection of the comprehensive validity assessment index set of historical data, and using ROC curves to select the optimal comprehensive validity assessment index threshold, this method is more automated and intelligent than traditional methods. It can dynamically adjust the validity judgment threshold based on real-time data, reducing manual intervention. For geomorphic anomaly data acquired in real time during gold mine exploration, this invention can extract key features in real time and compare them with the constructed comprehensive validity assessment model and thresholds to automatically identify and extract valid data, providing real-time support for exploration decisions. This invention, through intelligent analysis methods, can help exploration personnel more scientifically judge the validity of geophysical anomaly data, thereby effectively avoiding misjudgments and omissions, improving the scientificity and accuracy of gold mine exploration decisions, and reducing exploration costs and risks. Attached Figure Description

[0046] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0047] Figure 1 This is a schematic diagram of a geophysical anomaly data management system for gold mine exploration according to the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] Please see Figure 1 The present invention provides the following technical solution:

[0050] A geophysical anomaly data management system for gold exploration includes: a historical data management module, a feature extraction and analysis module, an effectiveness evaluation model construction module, and a real-time data acquisition and effectiveness evaluation module;

[0051] The historical data management module retrieves historical geomorphic anomaly data and historical event records from the database for areas where gold mine exploration tasks have been completed. Based on the historical event records, the historical geomorphic anomaly data is divided into historical valid data and historical invalid data.

[0052] The feature extraction and analysis module analyzes the deviation relationship between historical valid data and historical invalid data to identify the distinguishing features between the two, and marks the distinguishing features between historical valid data and historical invalid data as key validity features;

[0053] The effectiveness assessment model construction module constructs a comprehensive effectiveness assessment model for geomorphic anomaly data based on key effectiveness features and combining historical effective and invalid data, thereby obtaining a comprehensive effectiveness assessment index. Based on historical effective and invalid data, the corresponding comprehensive effectiveness assessment index is calculated, thereby obtaining the comprehensive effectiveness assessment threshold.

[0054] The real-time data acquisition and effectiveness assessment module collects real-time geophysical anomaly data of the gold mine exploration area and extracts key effectiveness features from the real-time geophysical anomaly data. Based on the key effectiveness features and combined with the effectiveness comprehensive assessment model, it calculates the real-time effectiveness comprehensive assessment index. The real-time effectiveness comprehensive assessment index is compared with the effectiveness comprehensive assessment threshold, and real-time effective data is extracted from the real-time geophysical anomaly data based on the comparison results.

[0055] The historical data management module includes a historical data acquisition unit and a historical data division and classification unit;

[0056] The historical data acquisition unit retrieves historical geomorphic anomaly data and corresponding historical event records from the database within the area where gold mine exploration tasks have been completed. The historical data division and classification unit judges the validity of geomorphic anomaly data based on historical event records, divides historical geomorphic anomaly data into valid data and invalid data, and constructs historical valid datasets and historical invalid datasets.

[0057] The feature extraction and analysis module includes a feature extraction and normalization unit and a feature difference analysis unit;

[0058] The feature extraction and normalization unit extracts features for each element in the historical valid dataset and the historical invalid dataset, and performs normalization processing to convert them into feature vectors; the feature difference analysis unit analyzes the feature differences between the historical valid dataset and the invalid dataset, thereby identifying the distinguishing features between the two and marking them as key features of validity.

[0059] The effectiveness evaluation model construction module includes a feature weight calculation unit and a comprehensive evaluation model construction unit;

[0060] The feature weight calculation unit calculates the weight coefficient of each key validity feature based on the differences between historical valid and invalid data; the comprehensive evaluation model construction unit uses the weight coefficients of each key validity feature to construct a comprehensive evaluation model for the validity of geomorphic anomaly data; the comprehensive validity evaluation index corresponding to historical geomorphic anomaly data is calculated through the comprehensive validity evaluation model, and the comprehensive validity evaluation threshold is determined.

[0061] The real-time data acquisition and effectiveness evaluation module includes a real-time data acquisition unit and a real-time effectiveness evaluation unit;

[0062] The real-time data acquisition unit collects real-time geophysical anomaly data of the gold mine exploration area and extracts key features of the real-time geophysical anomaly data to determine its validity. The real-time validity assessment unit calculates a comprehensive real-time validity assessment index based on the key features of the real-time data and compares it with a comprehensive validity assessment threshold to identify valid real-time data.

[0063] A method for managing geomorphic anomaly data in gold exploration includes the following steps:

[0064] Step S100. Obtain historical geophysical anomaly data and historical event records from the database within the area where the gold mine exploration task has been completed. Based on the historical event records, divide the historical geophysical anomaly data into historical valid data and historical invalid data.

[0065] Step S200. By analyzing the deviation relationship between historical valid data and historical invalid data, the distinguishing features between the two are identified, and the distinguishing features between historical valid data and historical invalid data are marked as key validity features;

[0066] Step S300. Based on the key features of effectiveness, and combining historical effective data and historical invalid data, construct a comprehensive evaluation model for the effectiveness of geomorphic anomaly data, thereby obtaining a comprehensive effectiveness evaluation index; calculate the corresponding comprehensive effectiveness evaluation index based on historical effective data and historical invalid data, thereby obtaining a comprehensive effectiveness evaluation threshold.

[0067] Step S400. Collect real-time geophysical anomaly data of the gold exploration area and extract key features of effectiveness from the real-time geophysical anomaly data; calculate the real-time effectiveness comprehensive evaluation index based on the key features of effectiveness and in combination with the effectiveness comprehensive evaluation model; compare the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold, and extract real-time effective data from the real-time geophysical anomaly data based on the comparison results.

[0068] Step S100 includes:

[0069] The historical event records refer to the records of historical geophysical anomalies within areas where gold exploration tasks have been completed. These records are used to determine the validity of historical geophysical anomaly data by comparing it with actual usage. Based on the historical event records, the historical geophysical anomaly data is divided into historically valid data and historically invalid data, and historically valid dataset A and historically invalid dataset B are constructed. Historically valid dataset A and historically invalid dataset B are represented as follows: A = {a1, a2, ..., an}, B = {b1, b2, ..., bm}, where A represents the historically valid dataset, a1 represents the first element of historically valid dataset A, a2 represents the second element of historically valid dataset A, and so on, with an representing the nth element of historically valid dataset A; B represents the historically invalid dataset, b1 represents the first element of historically invalid dataset B, b2 represents the second element of historically invalid dataset B, and so on, with bm representing the mth element of historically invalid dataset B.

[0070] Historical geophysical anomaly data refers to data detected during past geophysical, geochemical, and geological exploration processes that show significant differences from the normal background values ​​at the time. Specifically, geophysical anomaly data includes geological exploration anomaly data, geophysical anomaly data, and geochemical anomaly data. Geological exploration anomaly data refers to data discovered during geological exploration that is significantly different from the surrounding geological environment or expected geological features. This may include sudden changes in rock type, anomalies in stratigraphic structure, and abrupt changes in geological structures. These data often reveal the complexity of underground geological structures or the existence of special geological events. Geophysical anomaly data refers to data obtained during geophysical exploration by measuring geophysical fields (such as gravitational, magnetic, and electric fields). Geochemical anomaly data refers to data obtained through analysis of the content of elements or compounds in surface or underground samples during geochemical exploration. These data differ significantly from normal background values ​​and may manifest as gravity anomalies, magnetic anomalies, electrical anomalies, or abnormal changes in parameters such as seismic wave velocity and attenuation. They are usually related to the physical properties of underground geological bodies (such as density, magnetism, and electrical conductivity) and provide important information for inferring underground geological structures. Geochemical anomaly data refers to data that differs significantly from normal background values ​​and is found during geochemical exploration by analyzing the content of elements or compounds in surface or underground samples. These data may indicate the enrichment or depletion of certain elements or compounds and are usually related to the existence of underground ore bodies, geological tectonic activity, or migration of geochemical elements. They are important clues for mineral exploration and geological research.

[0071] In this embodiment, the specific numerical values ​​of geological exploration anomaly data, geophysical anomaly data, and geochemical anomaly data are presented as follows:

[0072] 1. Geological exploration anomaly data

[0073] Sudden changes in rock type: For example, during geological exploration, a region that was originally composed of dark metamorphic rocks may suddenly develop large amounts of light-colored igneous rocks. This type of anomaly can be detected through core sampling or surface outcrop investigation. Numerical representations may include:

[0074] Lithological changes: For example, the sudden appearance of granite or dikes in metamorphic rocks often indicates changes in geological structure or fault activity.

[0075] Anomalies in stratigraphic structure: If abnormal folds or faults are found in a certain layer, it may manifest as irregularities in the stratigraphic position and thickness.

[0076] 2. Geophysical Anomaly Data

[0077] Gravity anomalies: By measuring gravity values ​​at different underground levels, abnormal changes in the gravitational field can be detected. For example:

[0078] In a given region, the range of gravity values ​​may deviate from the usual 9.81 m / s² (normal gravitational acceleration of the Earth) to 9.90 m / s² or 9.75 m / s². These deviations may indicate an unusual distribution of underground material, such as high-density ore bodies or low-density rock bodies.

[0079] Magnetic Anomalies: When measuring the underground magnetic field with a magnetometer, certain areas may show abnormal magnetic field strength. For example:

[0080] The magnetic field strength of underground rock strata may deviate from the normal background value, such as 50 nT (nanotesla) in normal value, while anomalies may occur at 25 nT or less, which indicates the presence of weakly magnetic minerals or demagnetized fault zones in the area.

[0081] Electrical anomalies: Abnormal changes in underground resistivity detected through electrical resistivity surveys. For example:

[0082] The typical background resistivity is 20 Ω·m. If the resistivity of a certain area suddenly changes to 5 Ω·m or 100 Ω·m, it may indicate mineralization or different conductivity characteristics of the water source.

[0083] Seismic wave anomalies: During seismic exploration, the reflected wave velocity of underground geological layers changes, such as:

[0084] A sudden drop in P-wave velocity from the normal 6.5 km / s to 4.2 km / s may indicate the presence of softer sediments or fluid-filled porous structures underground.

[0085] 3. Geochemical Anomaly Data

[0086] Elemental enrichment anomalies: Analysis of elemental concentrations in subsurface samples through geochemical exploration reveals anomalous data that significantly differ from background values. For example:

[0087] The gold content in a normal background is 20 ppb, but in a certain area the gold concentration may reach 1000 ppb or higher, which is usually an indicator of ore body enrichment.

[0088] The background level for copper is 10 ppm, but in some mining areas, levels as high as 500 ppm may be found.

[0089] Anomalies in chemical composition: Abnormal changes in certain chemical components in soil or rock samples may indicate mineral deposits or geological activity. For example:

[0090] The sulfide concentration is about 0.1% in the background area, but can reach 5% or higher in the mineralized or tectonic zone, indicating that there may be relatively rich sulfide minerals in the area.

[0091] Step S200 includes:

[0092] S201. Based on historical valid dataset A and historical invalid dataset B, extract features from each element in historical valid dataset A and historical invalid dataset B respectively, and perform normalization processing, so that each element in historical valid dataset A and historical invalid dataset B is represented as a feature vector, and is represented as follows:

[0093] V_ai=[vai_1,vai_2,...,vai_p],V_bi=[vbi_1,vbi_2,...,vbi_p];

[0094] Where ai represents the i-th element in the historical valid dataset A, where i ranges from 1 to n; V_ai represents the feature vector corresponding to the i-th element in the historical valid dataset A, vai_1 represents the first-dimensional feature value of the feature vector V_ai, vai_2 represents the second-dimensional feature value of the feature vector V_ai, and so on, with vai_p representing the p-th dimension feature value of the feature vector V_ai, where p is the feature dimension, covering multiple indicators such as geology, geophysics, and geochemistry; similarly, V_bi represents the feature vector corresponding to the i-th element in the historical invalid dataset B, vbi_1 represents the first-dimensional feature value of the feature vector V_bi, vbi_2 represents the second-dimensional feature value of the feature vector V_bi, and so on, with vbi_p representing the p-th dimension feature value of the feature vector V_ai; due to the significant characteristics of geomorphic anomaly data such as high dimensionality and spatial distribution, commonly used algorithms include PCA, ICA, wavelet transform, autoencoder, SVM, K-means clustering, and neural networks. The specific algorithm chosen depends on the characteristics of the data (such as noise, dimensionality, spatial distribution, nonlinearity, etc.) and the requirements of the specific task. For example, if the data is noisy and requires dimensionality reduction, PCA or ICA may be more suitable; if it is necessary to capture multi-scale anomalies, wavelet transform may be more effective; for complex nonlinear relationships, neural networks and autoencoders may be better choices; the specific algorithm should be selected by the relevant personnel.

[0095] S202. For each element in the historical valid dataset A, calculate the difference between the feature values ​​of each dimension of the feature vectors V_ax and V_ay corresponding to two different elements ax and ay, thereby obtaining the minimum and maximum absolute values ​​of the difference between the feature values ​​of each dimension of the feature vectors in the historical valid dataset A, denoted as minΔvai_t and maxΔvai_t, respectively, where i takes values ​​from 1 to n and t takes values ​​from 1 to p; calculate the difference between the feature value of each dimension of the feature vector V_bi corresponding to each element in the historical invalid dataset B and the feature value of the corresponding dimension of V_ai corresponding to each element in the historical valid dataset A, thereby obtaining the absolute value of the difference Δvi_t; summarize the absolute values ​​of the difference Δvi_t of each dimension of the feature vector V_bi corresponding to each element in the historical invalid dataset B, thus forming the difference. Let Ct be a set, and Ct = {Δv1_t, Δv2_t, ..., Δvn_t}, where Δv1_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_a1 corresponding to the first element in the historical valid dataset A; Δv2_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_a2 corresponding to the second element in the historical valid dataset A; and so on, Δvn_t represents the absolute value of the difference between the t-th dimension eigenvalue of the feature vector V_bi corresponding to the element in the historical invalid dataset B and the t-th dimension eigenvalue of the feature vector V_an corresponding to the n-th element in the historical valid dataset A.

[0096] S203. For each element in the historical invalid dataset B, corresponding to the difference set Ct, compare the minimum and maximum absolute values ​​of the differences between each element in the difference set Ct and the feature values ​​of the corresponding dimension of the feature vector in the historical valid dataset A. Count the number N of elements in the difference set Ct that do not belong to the interval [minΔvai_t, maxΔvai_t]. If N / n ≥ d, then use the feature corresponding to the dimension of the difference set Ct as the distinguishing feature between the historical valid data and the historical invalid data, where d represents the proportion threshold. For each element in the historical invalid dataset B, corresponding to the difference set Ct, summarize the corresponding distinguishing features to form a distinguishing feature set Qt. Perform an intersection calculation on all distinguishing feature sets Qt, and mark the distinguishing features corresponding to the intersection calculation result as key validity features.

[0097] In this embodiment, the feature vectors corresponding to the elements in the historical valid dataset A and the historical invalid dataset B are represented as: V_ai=[vai_1,vai_2,...,vai_p], V_bi=[vbi_1,vbi_2,...,vbi_p], respectively; assuming the dimension p=4 of the feature vector, and the number of elements in the historical valid dataset A and the historical invalid dataset B are 4 and 3 respectively, then the correspondence between the elements in the historical valid dataset A and the feature vector V_ai is as follows:

[0098] a1→V_a1=[va1_1,va1_2,va1_3,va1_4];

[0099] a2→V_a2=[va2_1,va2_2,va2_3,va2_4];

[0100] a3→V_a3=[va3_1,va3_2,va3_3,va3_4];

[0101] a4→V_a4=[va4_1,va4_2,va4_3,va4_4];

[0102] Similarly, the correspondence between elements in the historical invalid dataset B and the feature vector V_bi is as follows:

[0103] b1→V_b1=[vb1_1,vb1_2,vb1_3,vb1_4];

[0104] b2→V_b2=[vb2_1,vb2_2,vb2_3,vb2_4];

[0105] b3→V_b3=[vb3_1,vb3_2,vb3_3,vb3_4];

[0106] For each element in the historical valid dataset A, the difference between the feature vectors V_ax and V_ay corresponding to two different elements ax and ay is calculated for each dimension. This yields the minimum and maximum absolute values ​​of the differences between the feature vectors for each dimension in the historical valid dataset A. For example, the calculation process for minΔva1_1 is as follows:

[0107] min{|va1_1-va2_1|,|va1_1-va3_1|,|va1_1-va4_1|,|va2_1-va3_1|,|va2_1-va4_1|,|va3_1-va4_1|}=minΔva1_1; Similarly, the calculation process for maxΔva1_1 is as follows:

[0108] max{|va1_1-va2_1|,|va1_1-va3_1|,|va1_1-va4_1|,|va2_1-va3_1|,|va2_1-va4_1|, |va3_1-va4_1|}=maxΔva1_1;

[0109] Taking feature vector V_b1 as an example, the feature value of each dimension of feature vector V_b1 is successively compared with the feature value of the corresponding dimension of the feature vector corresponding to each element in the historical valid dataset A to calculate the difference, thereby obtaining the absolute value of the difference. The absolute values ​​of the differences are summarized to form the difference set Ct; when t=1, the elements of the corresponding difference set C1 are:

[0110] |vb1_1-va1_1|,|vb1_1-va2_1|,|vb1_1-va3_1|,|vb1_1-va4_1|;

[0111] Each element in the difference set C1 is compared sequentially with the interval [minΔva1_1, maxΔva1_1] formed by the minimum and maximum absolute values ​​of the differences between the first dimension eigenvalues ​​of the feature vectors in the historical valid dataset A. Assuming that |vb1_1-va1_1|, |vb1_1-va2_1|, and |vb1_1-va3_1| do not belong to the interval [minΔva1_1, maxΔva1_1], then the number of elements in the difference set Ct that do not belong to the interval [minΔva1_t, maxΔva1_t] is N=3. Given N / n = 3 / 4 = 0.75, and assuming a ratio threshold d = 0.5, since N / n = 3 / 4 = 0.75 > d = 0.5, the feature corresponding to the first dimension feature value is taken as the distinguishing feature between historical valid data and historical invalid data. Traversing all dimension feature values ​​of feature vector V_b1, assuming the distinguishing features corresponding to feature vector V_b1 are: the feature corresponding to the first dimension feature value, the feature corresponding to the third dimension feature value, and the feature corresponding to the fourth dimension feature value, therefore, the distinguishing feature set Q1 corresponding to feature vector V_b1 is Q1 = {first dimension feature, third dimension feature, fourth dimension feature}.

[0112] Similarly, by analyzing eigenvectors V_b2 and V_b3 in the same way as eigenvector V_b1, we obtain the corresponding distinguishing feature sets Q2 and Q3. Assuming that distinguishing feature set Q1 ∩ distinguishing feature set Q2 ∩ distinguishing feature set Q3 = {first-dimensional feature, third-dimensional feature}, then the first-dimensional feature and the third-dimensional feature are taken as the key features of effectiveness.

[0113] Step S300 includes:

[0114] S301. For the key features of validity, search in the feature vectors corresponding to the historical valid data and historical invalid data to find the feature values ​​of the key features of validity in the feature vectors corresponding to the historical valid data and historical invalid data respectively, and calculate the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity, denoted as sk and mk respectively, where sk represents the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity k in the historical valid dataset A, and mk represents the average of the absolute values ​​of the differences of the feature values ​​of the corresponding dimensions of the key features of validity k in the historical invalid dataset B; calculate the weight coefficient wk of the key features of validity, and the specific calculation formula is: wk=|sk-mk| / σk, where σk represents the standard deviation of the feature values ​​corresponding to the key features of validity; traverse all key features of validity to obtain the weight coefficients corresponding to all key features of validity;

[0115] S302. Based on the weight coefficients corresponding to each key validity feature and combined with the feature values ​​corresponding to the key validity features, construct a comprehensive evaluation model for the validity of geomorphic anomaly data, and the corresponding calculation formula is: E=∑ R k=1 (wk×vk); where E represents the comprehensive effectiveness evaluation index, R represents the number of key effectiveness features, and vk represents the feature value corresponding to the k-th key effectiveness feature, and the feature value corresponding to the key effectiveness feature is found from the historical effective dataset A or the historical invalid dataset B; according to the comprehensive effectiveness evaluation model, the comprehensive effectiveness evaluation index E corresponding to each element in the historical effective dataset A and the historical invalid dataset B is calculated, and according to the division of the historical effective dataset A and the historical invalid dataset B, the comprehensive effectiveness evaluation index set EA of the historical effective dataset A and the comprehensive effectiveness evaluation index set EB of the historical invalid dataset B are obtained respectively. The comprehensive effectiveness evaluation index threshold E0 is obtained based on the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB. The comprehensive effectiveness evaluation index threshold E0 is used to determine whether the comprehensive effectiveness evaluation model needs to be adjusted, and the corresponding adjustments are made according to the judgment result.

[0116] In this embodiment, the feature vectors corresponding to the elements in the historical valid dataset A and the historical invalid dataset B are known to be represented as: V_ai=[vai_1,vai_2,...,vai_p], V_bi=[vbi_1,vbi_2,...,vbi_p], respectively. Assume the dimension p of the feature vectors is 4. And after analysis, the key features for validity are identified as the first and third dimensions. Therefore, the key feature values ​​for validity in historical valid dataset A are: vai_1 and vai_3, and the key feature values ​​for validity in historical valid dataset B are: vbi_1 and vbi_3. Thus, the feature values ​​corresponding to the key features for validity are uniformly represented as: v1 and v3.

[0117] Taking the eigenvalue v1 corresponding to the key validity feature as an example, calculate the average value of eigenvalue vai_1 in the historical valid dataset A, denoted as s1; calculate the average value of eigenvalue vbi_1 in the historical valid dataset B, denoted as m1; summarize the corresponding eigenvalues ​​vai_1 and vbi_1 in the historical valid datasets A and B, calculate the corresponding standard deviation, and obtain σ1; calculate the weight coefficient w1 of the key validity feature 1, and w1 = |s1 - m1| / σ1, assuming w1 = 0.6; perform the same analysis on the eigenvalue v3 corresponding to the key validity feature, assuming w3 = 0.2; then according to the formula: E = ∑ R k=1 (wk×vk) Calculate the comprehensive validity evaluation index for each element in the historical valid dataset A and the historical invalid dataset B respectively. The comprehensive validity evaluation index for each element in the historical valid dataset A is E=0.6×vai_1+0.4×vai_3, and the comprehensive validity evaluation index for each element in the historical valid dataset B is E=0.6×vbi_1+0.4×vbi_3. Thus, we obtain the comprehensive validity evaluation index set EA and the comprehensive validity evaluation index set EB.

[0118] The values ​​corresponding to the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB are represented on a number axis. Based on the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB, the threshold value E0 of the comprehensive effectiveness evaluation index is obtained. The specific analysis is as follows:

[0119] Calculate the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB. If the intersection result is an empty set, extract the minimum value EA_min and the maximum value EA_max from the comprehensive effectiveness evaluation index set EA, and the minimum value EB_min and the maximum value EB_max from the comprehensive effectiveness evaluation index set EB. If EA_min > EB_max, then EA_min is used as the comprehensive effectiveness evaluation index threshold E0; if EA_max < EB_min, then EA_max is used as the comprehensive effectiveness evaluation index threshold E0.

[0120] If the intersection calculation result is not an empty set, compare the maximum value EA_max in the comprehensive effectiveness evaluation index set EA with the maximum value EB_max in the comprehensive effectiveness evaluation index set EB. If EA_max ≤ EB_max, then the elements in the intersection region are successively used as the undetermined comprehensive effectiveness evaluation index threshold E1. Under the undetermined comprehensive effectiveness evaluation index threshold E1, the values ​​corresponding to true positive TP, false positive FP, true negative TN, and false negative FN are obtained in sequence. Among them, true positive TP represents the number of samples that were successfully and correctly classified as comprehensive effectiveness evaluation index within the intersection region; false positive FP represents the number of samples that were misclassified as comprehensive effectiveness evaluation index within the intersection region; and true negative TN represents the number of samples that were successfully and correctly classified as comprehensive ineffectiveness evaluation index outside the intersection region. False negatives (FN) represent the number of samples misjudged as invalidity comprehensive assessment index outside the intersection region; calculate the true positive rate (TPR) and false positive rate (FPR), where the specific calculation formulas are: TPR = TP / (TP + FN), FPR = FP / (FP + TN); summarize the true positive rate (TPR) and false positive rate (FPR) under each undetermined validity comprehensive assessment index threshold E1, and plot the ROC curve, with the horizontal axis of the ROC curve being FPR and the vertical axis being TPR; select the undetermined validity comprehensive assessment index threshold E1 corresponding to the point in the ROC curve that is farthest from the (0,1) point as the validity comprehensive assessment index threshold E0; similarly, if EB_max ≤ EA_max, analyze according to the analysis process for the case of EA_max ≤ EB_max, thereby obtaining the validity comprehensive assessment index threshold E0;

[0121] The effectiveness comprehensive evaluation model is determined based on the effectiveness comprehensive evaluation index threshold E0 to determine whether it needs adjustment, and corresponding adjustments are made based on the determination results, as detailed below:

[0122] If the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is an empty set, then the current comprehensive effectiveness evaluation model does not need to be adjusted. If the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not an empty set, and the intersection calculation results are all proper subsets of the comprehensive effectiveness evaluation index sets EA and EB, then the current comprehensive effectiveness evaluation model needs to be adjusted. The specific adjustment process is as follows:

[0123] An adaptive adjustment mechanism is used to dynamically adjust the weight coefficients of key effectiveness features. The specific calculation formula is: w'k = αk × wk + βk, where αk and βk represent the adjustment factor and bias term corresponding to key effectiveness feature k, respectively. The weight coefficients of all key effectiveness features are iterated to obtain several adjusted weight coefficients, which are then normalized to ensure that the sum of the weight coefficients of all adjusted key effectiveness features equals 1. The expression for the adjusted comprehensive effectiveness evaluation model is: E_end = ∑ R k=1 (w'k×vk); Based on the adjusted comprehensive effectiveness evaluation model, calculate the comprehensive effectiveness evaluation index E_end for each element in the historical valid dataset A and the historical invalid dataset B, thereby obtaining the corresponding comprehensive effectiveness evaluation index set EA_end and comprehensive effectiveness evaluation index set EB_end. The intervals represented by the adjusted comprehensive effectiveness evaluation index set EA_end and comprehensive effectiveness evaluation index set EB_end on the number axis do not intersect, thus obtaining the corresponding comprehensive effectiveness evaluation index threshold E0.

[0124] In this embodiment, the values ​​corresponding to the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB are represented on a number axis. The intersection of the comprehensive effectiveness evaluation index sets EA and EB is calculated. Assuming the interval corresponding to the comprehensive effectiveness evaluation index set EA on the number axis is [1,5] and the interval corresponding to the comprehensive effectiveness evaluation index set EB is [4,10], the intersection interval is [4,5]. Therefore, the elements within the intersection region [4,5] are used as the threshold E1 of the undetermined comprehensive effectiveness evaluation index. Under the threshold of E1, the values ​​corresponding to true positive (TP), false positive (FP), true negative (TN), and false negative (FN) are obtained sequentially, and the true positive rate (TPR) and false positive rate (FPR) are calculated. The true positive rate (TPR) and false positive rate (FPR) under each undetermined comprehensive effectiveness assessment index threshold of E1 are summarized, and an ROC curve is plotted, with the horizontal axis of the ROC curve being FPR and the vertical axis being TPR. The undetermined comprehensive effectiveness assessment index threshold of E1 corresponding to the point in the ROC curve that is farthest from the (0,1) point is selected as the comprehensive effectiveness assessment index threshold of E0, assuming that the comprehensive effectiveness assessment index threshold of E0 = 5.

[0125] Because when the validity comprehensive evaluation index threshold E0=5, elements in historical invalid datasets may be misclassified as historical valid data, the current validity comprehensive evaluation model needs parameter adjustment. An adaptive adjustment mechanism is used to dynamically adjust the weight coefficients of key validity features. The specific calculation formula is: w'k=αk×wk+βk. Assuming the adjusted weight coefficients of key validity features are 0.3 and 0.7 respectively, the expression of the adjusted validity comprehensive evaluation model is: E_end=0.3×vbi_1+0.7×vbi_3. According to the calculation method that the intersection of the validity comprehensive evaluation index set EA and the validity comprehensive evaluation index set EB is an empty set, the corresponding validity comprehensive evaluation index threshold E0 is obtained. Assuming E0=4.5, and according to the adjusted validity comprehensive evaluation model, the interval corresponding to the validity comprehensive evaluation index set EA is [0,4.5], and the interval corresponding to the validity comprehensive evaluation index set EB is (4.5,7).

[0126] The specific analysis process for comparing the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold in step S400, and extracting real-time effective data from the real-time geophysical anomaly data based on the comparison results, is as follows:

[0127] When the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is empty, and EA_min > EB_max, then EA_min is used as the comprehensive effectiveness evaluation index threshold E0; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0, and when E' ≥ E0, the real-time materialized abnormal data is considered real-time valid data; when the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is empty, and EA_max < EB_min, then EA_max is used as the comprehensive effectiveness evaluation index threshold E0; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0, and when E' ≤ E0, the real-time materialized abnormal data is considered real-time valid data.

[0128] When the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not empty, and EA_max≤EB_max; the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0. When E'≤E0, the real-time materialized abnormal data is considered real-time valid data; when the intersection of the comprehensive effectiveness evaluation index set EA and the comprehensive effectiveness evaluation index set EB is not empty, and EB_max≤EA_max, the real-time comprehensive effectiveness evaluation index E' is compared with the comprehensive effectiveness evaluation index threshold E0. When E'≥E0, the real-time materialized abnormal data is considered real-time valid data.

[0129] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0130] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for managing geochemical anomaly data for gold exploration, characterized in that: The method comprises the following steps: Step S100. Obtain historical geochemical anomaly data and historical event records in the completed gold exploration task area from the database, and divide the historical geochemical anomaly data into historical effective data and historical invalid data according to the historical event records; Step S200. Identify the distinguishing features between the historical effective data and the historical invalid data by analyzing the deviation relationship between the historical effective data and the historical invalid data, and mark the distinguishing features between the historical effective data and the historical invalid data as effectiveness key features; Step S300. Based on the effectiveness key features, combine the historical effective data and the historical invalid data to construct an effectiveness comprehensive evaluation model of the geochemical anomaly data, and obtain an effectiveness comprehensive evaluation index; calculate the corresponding effectiveness comprehensive evaluation index according to the historical effective data and the historical invalid data, and obtain an effectiveness comprehensive evaluation threshold; specifically: Calculate the average value sk and mk of the absolute value of the difference of the feature value corresponding to the effectiveness key feature k in the historical effective data and the historical invalid data, and the standard deviation σk of the feature value corresponding to the effectiveness key feature k, and the weight coefficient wk, the formula is: wk=|sk-mk| / σk; An effective comprehensive evaluation model of geophysical anomaly data is constructed: E=∑ R k=1 (wk×vk), wherein E is an effective comprehensive evaluation index, R is the number of effective key features, vk is the feature value corresponding to the kth effective key feature; the effective comprehensive evaluation index is calculated, and the effective comprehensive evaluation index set EA of the historical effective data set and the effective comprehensive evaluation index set EB of the historical invalid data set are formed; Obtain the effectiveness comprehensive evaluation index threshold E0 according to EA and EB, specifically: calculate the intersection of EA and EB, if the intersection is empty, extract the minimum value EA_min and the maximum value EA_max of EA, and the minimum value EB_min and the maximum value EB_max of EB; if EA_min>EB_max, then EA_min is taken as E0; if EA_max<EB_min, then EA_max is taken as E0; If the intersection is not empty, compare the size relationship of EA_max and EB_max, if EA_max≤EB_max, then the elements in the intersection region are taken as the to-be-determined threshold E1 in turn, the true positive rate TPR and the false positive rate FPR are calculated, the ROC curve is drawn, and the E1 corresponding to the point farthest from the point (0, 1) is selected as E0; if EB_max≤EA_max, then E0 is obtained according to the EA_max≤EB_max analysis logic; According to the E0 judgment validity comprehensive evaluation model whether adjustment, specific for: EA and EB no intersection when validity comprehensive evaluation model need not adjust, have intersection then according to w'k=αk×wk+βk dynamic adjustment validity key feature weight coefficient and normalization, wherein αk and βk respectively validity key feature k corresponding adjustment factor and bias term; Build adjusted validity comprehensive evaluation model for: E_end=∑ R k=1 (w'k×vk), calculate validity comprehensive evaluation index, get EA_end, EB_end, and ensure that there is no intersection and determine the corresponding E0; Step S400. Collect real-time geochemical anomaly data of the gold exploration area, extract the effectiveness key features from the real-time geochemical anomaly data; calculate the real-time effectiveness comprehensive evaluation index according to the effectiveness key features and combining the effectiveness comprehensive evaluation model; compare the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold, and extract real-time effective data from the real-time geochemical anomaly data according to the comparison result.

2. The method according to claim 1, wherein: The step S100 comprises: The historical event record refers to record information for judging whether historical geochemical anomaly data in a completed gold exploration task area is effective by comparing the use of actual historical geochemical anomaly data; according to the historical event record, the historical geochemical anomaly data is divided into historical effective data and historical ineffective data, and a historical effective data set A and a historical ineffective data set B are constructed, wherein the historical effective data set A and the historical ineffective data set B are respectively represented as: A={a1, a2,..., an}, B={b1, b2,..., bm}, wherein A represents the historical effective data set, a1 represents the first element in the historical effective data set A, a2 represents the second element in the historical effective data set A, and so on, and an represents the nth element in the historical effective data set A; B represents the historical ineffective data set, b1 represents the first element in the historical ineffective data set B, b2 represents the second element in the historical ineffective data set B, and so on, and bm represents the mth element in the historical ineffective data set B.

3. The method according to claim 2, wherein: The step S200 comprises: S201. Based on the historical effective data set A and the historical ineffective data set B, feature extraction is performed on each element in the historical effective data set A and the historical ineffective data set B respectively, and normalization processing is performed, so that each element in the historical effective data set A and the historical ineffective data set B is respectively represented as a feature vector, and is respectively represented as: V_ai=[vai_1,vai_2,...,vai_p],V_bi=[vbi_1,vbi_2,...,vbi_p]; Wherein, ai represents the ith element in the historical effective data set A, i takes 1 to n; V_ai represents the feature vector corresponding to the ith element in the historical effective data set A, vai_1 represents the first dimensional feature value of the feature vector V_ai, vai_2 represents the second dimensional feature value of the feature vector V_ai, and so on, and vai_p represents the pth dimensional feature value of the feature vector V_ai, and p is the feature dimension; similarly, V_bi represents the feature vector corresponding to the ith element in the historical ineffective data set B, vbi_1 represents the first dimensional feature value of the feature vector V_bi, vbi_2 represents the second dimensional feature value of the feature vector V_bi, and so on, and vbi_p represents the pth dimensional feature value of the feature vector V_ai. S202. For each element in the historical valid data set A, the feature values of each dimension of the feature vectors V ax and V ay corresponding to the different two elements ax and ay are sequentially calculated by difference, so as to obtain the minimum and maximum of the absolute value of the difference of the feature values of each dimension of the feature vectors in the historical valid data set A, denoted as minΔvai_t and maxΔvai_t respectively, wherein i takes 1 to n, and t takes 1 to p; the feature values of each dimension of the feature vector V_bi corresponding to each element in the historical invalid data set B are sequentially calculated by difference with the feature values of the corresponding dimension of V_ai corresponding to each element in the historical valid data set A, so as to obtain the absolute value of the difference Δvi_t; the absolute values of the difference Δvi_t of each dimension of the feature vector V_bi corresponding to each element in the historical invalid data set B are summarized, so as to constitute the difference set Ct, and Ct={Δv1_t,Δv2_t,...,Δvn_t}, wherein Δv1_t represents the absolute value of the difference between the feature value of the tth dimension in the feature vector V_bi corresponding to the element in the historical invalid data set B and the feature value of the tth dimension in the feature vector V_a1 corresponding to the first element in the historical valid data set A, Δv2_t represents the absolute value of the difference between the feature value of the tth dimension in the feature vector V_bi corresponding to the element in the historical invalid data set B and the feature value of the tth dimension in the feature vector V_a2 corresponding to the second element in the historical valid data set A, and so on, and Δvn_t represents the absolute value of the difference between the feature value of the tth dimension in the feature vector V_bi corresponding to the element in the historical invalid data set B and the feature value of the tth dimension in the feature vector V_an corresponding to the n th element in the historical valid data set A; S203. For each element in the historical invalid data set B, the minimum minΔvai_t and the maximum maxΔvai_t of the absolute value of the difference of the feature values of the corresponding dimension of the feature vector in the historical valid data set A are compared with each element in the difference set Ct, the number N of elements in the difference set Ct that do not belong to the interval [minΔvai_t, maxΔvai_t] is counted, and if N / n≥d, the feature corresponding to the dimension of the difference set Ct is taken as the distinguishing feature between the historical valid data and the historical invalid data, wherein d represents the proportion threshold; for each element in the historical invalid data set B, the corresponding distinguishing feature is summarized to constitute the distinguishing feature set Qt; the intersection of all the distinguishing feature sets Qt is calculated, and the distinguishing feature corresponding to the intersection calculation result is marked as the validity key feature.

4. The method of claim 1, wherein: The specific analysis process of comparing the real-time validity comprehensive evaluation index with the validity comprehensive evaluation threshold in step S400 and extracting real-time valid data from real-time geochemical anomaly data according to the comparison result is as follows: When the intersection of the effective comprehensive evaluation index set EA and the effective comprehensive evaluation index set EB is empty, and EA_min>EB_max, EA_min is taken as the effective comprehensive evaluation index threshold E0; the real-time effective comprehensive evaluation index E' is compared with the effective comprehensive evaluation index threshold E0, when E'≥E0, the real-time geophysical and geochemical anomaly data is real-time effective data; when the intersection of the effective comprehensive evaluation index set EA and the effective comprehensive evaluation index set EB is empty, and EA_max<EB_min, EA_max is taken as the effective comprehensive evaluation index threshold E0; the real-time effective comprehensive evaluation index E' is compared with the effective comprehensive evaluation index threshold E0, when E'≤E0, the real-time geophysical and geochemical anomaly data is real-time effective data; When the intersection of the effective comprehensive evaluation index set EA and the effective comprehensive evaluation index set EB is not empty, and EA_max≤EB_max; the real-time effective comprehensive evaluation index E' is compared with the effective comprehensive evaluation index threshold E0, when E'≤E0, the real-time geophysical and geochemical anomaly data is real-time effective data; when the intersection of the effective comprehensive evaluation index set EA and the effective comprehensive evaluation index set EB is not empty, and EB_max≤EA_max, the real-time effective comprehensive evaluation index E' is compared with the effective comprehensive evaluation index threshold E0, when E'≥E0, the real-time geophysical and geochemical anomaly data is real-time effective data.

5. A system for managing geochemical anomaly data for gold exploration, applied to the method for managing geochemical anomaly data for gold exploration according to any one of claims 1-4, characterized in that: The system comprises a historical data management module, a feature extraction and analysis module, an effectiveness evaluation model construction module and a real-time data acquisition and effectiveness evaluation module; The historical data management module obtains historical geophysical and geochemical anomaly data and historical event records in a completed gold exploration task area from a database, and divides the historical geophysical and geochemical anomaly data into historical effective data and historical ineffective data according to the historical event records; The feature extraction and analysis module analyzes the deviation relationship between the historical effective data and the historical ineffective data, thereby identifying the distinguishing features between the two, and marking the distinguishing features between the historical effective data and the historical ineffective data as effectiveness key features; The effectiveness evaluation model construction module constructs an effectiveness comprehensive evaluation model of geophysical and geochemical anomaly data based on the effectiveness key features and in combination with the historical effective data and the historical ineffective data, thereby obtaining an effectiveness comprehensive evaluation index; and calculates the corresponding effectiveness comprehensive evaluation index according to the historical effective data and the historical ineffective data, thereby obtaining an effectiveness comprehensive evaluation threshold; The real-time data acquisition and effectiveness evaluation module acquires real-time geophysical and geochemical anomaly data of a gold exploration area, extracts effectiveness key features from the real-time geophysical and geochemical anomaly data; calculates a real-time effectiveness comprehensive evaluation index according to the effectiveness key features and in combination with the effectiveness comprehensive evaluation model; compares the real-time effectiveness comprehensive evaluation index with the effectiveness comprehensive evaluation threshold, and extracts real-time effective data from the real-time geophysical and geochemical anomaly data according to the comparison result.

6. The geochemical anomaly data management system for gold mine exploration according to claim 5, characterized in that: The historical data management module comprises a historical data acquisition unit and a historical data division and classification unit; The historical data acquisition unit acquires historical geochemical anomaly data and corresponding historical event records in a completed gold exploration task area from a database; the historical data division and classification unit judges the validity of the geochemical anomaly data according to the historical event records, divides the historical geochemical anomaly data into valid data and invalid data, and constructs a historical valid data set and a historical invalid data set.

7. The geochemical anomaly data management system for gold mine exploration according to claim 5, characterized in that: The feature extraction and analysis module comprises a feature extraction and normalization unit and a feature difference analysis unit; The feature extraction and normalization unit extracts each feature and performs normalization processing for each element in the historical valid data set and the historical invalid data set, thereby converting into a feature vector; The feature difference analysis unit analyzes the feature difference between the historical valid data set and the invalid data set, thereby identifying the distinguishing features therebetween and marking as validity key features.

8. The geochemical anomaly data management system for gold mine exploration according to claim 5, characterized in that: The validity evaluation model construction module comprises a feature weight calculation unit and a comprehensive evaluation model construction unit; The feature weight calculation unit calculates the weight coefficient of each validity key feature based on the difference of the validity key features in the historical valid data and invalid data; the comprehensive evaluation model construction unit constructs a validity comprehensive evaluation model of the geochemical anomaly data by using the weight coefficients of the validity key features; The validity comprehensive evaluation model is used to calculate the validity comprehensive evaluation index corresponding to the historical geochemical anomaly data, and a validity comprehensive evaluation threshold is determined; The real-time data acquisition and validity evaluation module comprises a real-time data acquisition unit and a real-time validity evaluation unit; the real-time data acquisition unit acquires real-time geochemical anomaly data of a gold exploration area and extracts validity key features of the real-time geochemical anomaly data; the real-time validity evaluation unit calculates a real-time validity comprehensive evaluation index according to the validity key features of the real-time data, and compares the real-time validity comprehensive evaluation index with the validity comprehensive evaluation threshold, thereby identifying real-time valid data.

Images