Geological exploration data analysis system and method based on multi-source data fusion

Abstract

The invention discloses a geological exploration data analysis system and method based on multi-source data fusion, and relates to the technical field of data analys.The geological exploration data analysis method comprises the steps that a multi-source sensor is used for collecting general characteristic data in geology, and real-time characteristic values are calculated to construct an environment characteristic matrix; calculating the similarity between the real-time environment characteristic matrix and each type of geological data in the knowledge base, and selecting and judging to obtain a real-time survey geological type; respectively calculating data fusion weights of the real-time survey geological types by using the quality score and the relevancy; performing weighted fusion on the real-time data flow during geological survey by using the calculated data fusion weight; according to different types of geological historical survey data, abnormal thresholds of different types of geology are calculated and standardized, and an abnormal threshold database is generated; and extracting an abnormal threshold of the type of the real-time surveyed geology from the abnormal threshold database, judging the fused data by using the abnormal threshold, obtaining whether the real-time surveyed geology is abnormal or not, and performing early warning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a geological exploration data analysis system and method based on multi-source data fusion. BACKGROUND

[0002] Geological exploration aims to comprehensively understand the geological structure of the earth's surface and interior, the distribution of mineral resources and other information. However, the information provided by a single data source such as geological mapping, geophysical exploration, geochemical exploration, remote sensing, etc. is limited and cannot fully depict complex geological phenomena and accurately predict the location of mineral resources. Multi-source data fusion technology can integrate data from different sources, take advantage of each source, and improve exploration efficiency and accuracy. With the development of computer technology, computers have been applied to the field of geology, mainly for processing geophysical and geochemical data such as simple data filtering, statistical analysis, etc. Initially, different types of geophysical data (such as gravity and magnetic data) were integrated and analyzed to infer the underground geological structure, but the fusion method was relatively simple, mainly based on intuitive comparison and manual drawing of integrated maps. GIS technology gradually matured and was widely used in geological exploration, providing a unified spatial analysis platform for multi-source data fusion. With the continuous progress of information technology, a variety of professional geological exploration data processing and analysis systems have emerged and gradually developed towards integration.

[0003] However, there are large differences in geological environments in the earth, and there are some special geological environments. In today's geological survey data analysis system, ordinary analysis methods and judgment standards suitable for most common geology are usually used; for special geological analysis, the general analysis system cannot give accurate analysis and judgment results, and it is crucial to change the analysis and judgment standards according to different geological types and reduce judgment errors. SUMMARY

[0004] The present application relates to the technical field of data analysis, in particular to a geological exploration data analysis system and method based on multi-source data fusion.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: The geological exploration data analysis method based on multi-source data fusion comprises the following steps: S100, screening environmental data in all types of geology, extracting environmental data that exist in all types of geology and obtaining general characteristics through professional knowledge review, collecting general characteristic data in geology by using multi-source sensors, calculating real-time characteristic values to construct an environmental characteristic matrix; Further, the specific steps of calculating real-time characteristic values to construct an environmental characteristic matrix are: S101, collect all types of geological environmental data, find the types of environmental data that exist in all types of geology, use expert knowledge to review and screen the types of environmental data that exist in all types of geology, and obtain the general characteristics of all types of geology; S102, when real-time surveying geology, use sensors to collect general characteristic data of real-time surveying geology, calculate and analyze the general characteristic data to obtain characteristic values, the formula is: ; In the formula, e i represents the calculated i-th characteristic value, f i represents the i-th characteristic extraction function, which includes but is not limited to average slope, water depth; Sen represents the general characteristic data; For each general characteristic data, calculate the characteristic value and standardize it; S103, use all the calculated characteristic values to construct an environmental characteristic matrix E = [e1, e2, e3, …, en] n ] T , e1, e2, e3, …, en n represents the 1st, 2nd, 3rd, …, nth characteristic value, n is a positive integer.

[0006] By screening the common environmental data of all types of geology (such as soil humidity, rock stress, geomagnetic intensity, etc.), the data fragmentation caused by the difference of geological types is avoided, and a unified foundation is laid for subsequent cross-type analysis. Constructing an environmental characteristic matrix facilitates the rapid processing of massive real-time data by computers and provides structured data support for subsequent real-time analysis.

[0007] S200, according to professional knowledge, predefine the general characteristic range of each type of geology to generate a knowledge base; calculate the similarity between the real-time environmental characteristic matrix and each type of geological data in the knowledge base, and select and judge to obtain the real-time surveying geological type; Further, the specific steps for selecting and judging the real-time surveying geological type are as follows: S201, according to professional knowledge, predefine the general characteristic range of each type of geology, construct a geological database containing all types, generate a type label g for each type of geology in the geological database, calculate the average value and standard deviation of different general data for each type of geology in the geological database, calculate the similarity between the real-time environmental characteristic matrix and each type of geology in the geological database, and the formula is: ; In the formula, Sim k represents the similarity between the real-time environmental characteristic matrix and the k-th type of geology in the geological database, e i represents the i-th characteristic value in the real-time environmental characteristic matrix, μki represents the average value of the i-th general feature in the k-th type of geology, sd ki represents the standard deviation of the i-th general feature in the k-th type of geology; S202, a similarity screening mechanism is constructed, and the formula is: In the formula, g t represents the type label of the real-time survey geology, and the formula represents the meaning of selecting k from all types of geology k to make Sim k The type label corresponding to k which makes the maximum value is taken as the type label g t of the real-time survey geology. The similarity screening mechanism is used to judge and screen the similarity of the real-time environment feature matrix and all types of geology in the geological database, and the type label of the real-time survey geology is obtained.

[0008] By matching real-time data with the knowledge base through similarity calculation, complex geological types can be automatically identified, and subjective errors of manual interpretation can be avoided. The geological type is determined in advance, which provides a "type label" for subsequent data fusion and anomaly judgment, and solves the defects of "one-size-fits-all" analysis of general systems.

[0009] S300, the quality score and the correlation of the real-time general feature data are calculated, and the data fusion weight of the real-time survey geology type is calculated respectively by using the calculated quality score and the correlation; Further, the specific steps of calculating the data fusion weight of the real-time survey geology type by using the calculated quality score and the correlation are: S301, for the real-time survey geology type label, expert scoring is performed on each general feature in the real-time survey geology type by using expert knowledge, the historical accuracy of each general feature is calculated, and the formula is: In the formula, Lz represents the historical accuracy of the general feature, Cr represents the number of correct detections in the history of the general feature, and Cz represents the total number of uses of the general feature in the history; and the average value of the expert score and the historical accuracy of each general feature is taken as the correlation of the general feature. The signal-to-noise ratio, the theoretical maximum signal-to-noise ratio and the data coverage of each general feature in the historical data source are collected, the quality score of each general feature is calculated, and the formula is: In the formula, q represents the quality score of the general feature, SNR represents the signal-to-noise ratio of the general feature in the historical data source, SNR max represents the theoretical maximum signal-to-noise ratio in the historical data source, and Cover represents the coverage of the general feature. S302, the fusion weight is calculated by using the correlation and the quality score of each general feature, and the formula is:​​ ; In the formula, w j represents the fusion weight of the jth general feature, Rel j represents the relevance of the jth general feature, q j represents the quality score of the jth general feature; m represents the total number of general features, and the fusion weight of all general features is obtained by repeated calculation.

[0010] In combination with the quality score and the relevance, noise data (such as abnormal values caused by sensor failure) is removed, and data reliability is improved. Through dynamic weight calculation, the weight of an anti-interference sensor (such as a fiber strain sensor) is automatically increased for a special geological environment (such as a high magnetic interference area), the influence of interference data (such as a traditional electromagnetic sensor) is reduced, and the adaptability of the system is enhanced.

[0011] S400, in real-time geological survey, using a sensor to collect environmental data to generate a real-time data stream, and using the calculated data fusion weight to weight and fuse the real-time data stream in the geological survey; Further, the specific steps of using the calculated data fusion weight to weight and fuse the real-time data stream in the geological survey are: S401, in real-time geological survey, using a sensor to collect environmental data to generate a real-time data stream, extracting real-time general feature data values in the real-time data stream, and using the fusion weight to weight and fuse, the formula is: ; In the formula, Rt represents the fused data after weight fusion, Sd j represents the real-time data value of the jth general feature.

[0012] Weighted fusion of real-time data stream can weaken sudden interference in real time and ensure data continuity. Fusion of multi-dimensional data such as geology, hydrology and meteorology forms a "three-dimensional geological portrait" and solves the one-sidedness of single data dimension analysis. Through weight adjustment, data real-time is ensured (such as second-level update), and error accumulation caused by rapid acquisition is avoided.

[0013] S500, according to different types of geological historical survey data, calculating the abnormal threshold of different types of geology and standardizing to generate an abnormal threshold database; Further, the specific steps of generating the abnormal threshold database are: S501, for each type of geology, extracting the feature values in the environmental feature matrix in the history and standardizing, calculating the mean and standard deviation of the historical feature values, and using the feature values to calculate the environmental variation coefficient of each type of geology, the formula is: ; In the formula, EnvVar represents the environmental variation coefficient of each type of geology, e i represents the i-th feature value in the real-time environmental feature matrix, μ i represents the average value of the i-th historical feature value, sd i represents the standard deviation of the i-th historical feature value; S502, extract the average value and standard deviation of the fused data of each type of geology in history, and standardize, calculate the abnormal threshold of each type of geology, the formula is: ; In the formula, τ k represents the abnormal threshold of the k-th type of geology, Rμ k represents the average value of the historical fused data of the k-th type of geology, Rsd k represents the standard deviation of the historical fused data of the k-th type of geology, and α represents the sensitivity coefficient; An abnormal threshold database is constructed using the abnormal thresholds of all types of geology.

[0014] Based on the historical survey data of different types of geology, a more practical scene-based abnormal standard is established to avoid misjudgment caused by a general threshold.

[0015] S600, extract the abnormal threshold of the real-time survey geology type in the abnormal threshold database, and use the abnormal threshold to judge the fused data to obtain whether the real-time survey geology exists abnormal and give an early warning.

[0016] Further, the specific steps of obtaining whether the real-time survey geology exists abnormal and giving an early warning are: S601, extract the abnormal threshold τ gt of the real-time survey geology type in the abnormal threshold database, use the abnormal threshold of the real-time survey geology type to judge the fused data of the real-time survey geology, when Rt> τ gt , it is judged that the real-time survey geology exists abnormal risk and gives an early warning; when Rt≤ τ gt , it is judged that the real-time survey geology does not exist abnormal risk.

[0017] Based on the real-time geology type, the corresponding threshold is extracted (such as the development threshold of karst cave of karst landform and the landslide threshold of loess plateau), which reduces the false negatives caused by the "unified standard" of the general system. Combined with the fused data and the threshold, multi-level early warning (such as yellow warning→red warning) can be realized to provide a time window for emergency decision-making.

[0018] The geological survey data analysis system based on multi-source data fusion includes a data acquisition module, a feature analysis module, a survey type judgment module, a fusion module, an abnormal threshold module, and an abnormal judgment module. The data collection module is used for collecting environmental data of all types of geology, and extracting feature values in the environmental feature matrix in history; The feature analysis module is used for extracting general features of all types of geology, and calculating and analyzing general feature data to obtain feature values; The survey type judgment module is used for calculating the similarity between the real-time feature value and each type of geology in the geological database, and screening the maximum value to obtain the real-time survey geological type; The fusion module is used for calculating the fusion weight of each general feature, and weighting and fusing the general features in the real-time data stream by using the fusion weight; The anomaly threshold module is used for calculating the anomaly threshold by using the average value, standard deviation and environmental variation coefficient of the historical fusion data; The anomaly judgment module is used for judging the fusion data of the real-time survey geology by using the anomaly threshold, and obtaining whether the real-time survey geology exists anomaly and warning.

[0019] The feature analysis module includes a general feature unit and a feature value unit; The general feature unit is used for reviewing and screening the types of environmental data existing in all types of geology by using expert knowledge, to obtain the general features of all types of geology; The feature value unit is used for collecting general feature data of the real-time survey geology by using a sensor, and calculating and analyzing the general feature data to obtain feature values.

[0020] The fusion module includes a fusion weight unit and a data fusion unit; The fusion weight unit is used for calculating the correlation degree and quality score of each general feature, and calculating the fusion weight of each general feature; The data fusion unit is used for collecting environmental data by using a sensor to generate a real-time data stream, extracting real-time general feature data values in the real-time data stream, and weighting and fusing by using the fusion weight.

[0021] Compared with the prior art, the beneficial effects of the present application are: 1、The present application realizes "one strategy for one type of geology" analysis through the judgment of the real-time survey geological type and the calculation of the anomaly threshold respectively; different judgment standards are adopted for different types of geology, so that the geological judgment is more accurate and misjudgment is avoided.

[0022] 2、The present application weakens the interference influence of a single sensor through the fusion of general features in high magnetic, high seismic and other special scenes (such as volcanic activity area) by using multi-source sensor weighted fusion. BRIEF DESCRIPTION OF DRAWINGS

[0023] Fig. 1 It is a module distribution diagram of the geological exploration data analysis system based on multi-source data fusion of the present application; Fig. 2 This is a schematic diagram illustrating the steps of the geological exploration data analysis method based on multi-source data fusion according to the present invention. Detailed Implementation

[0024] 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.

[0025] Example: Figs. 1-2 As shown, the present invention provides a technical solution. A geological exploration data analysis method based on multi-source data fusion, the method comprising the following steps: S100. Screen environmental data from all types of geology, extract all existing environmental data and obtain common features through professional knowledge review, collect common feature data from geology using multi-source sensors, calculate real-time feature values ​​and construct an environmental feature matrix. The specific steps for calculating real-time feature values ​​and constructing the environmental feature matrix are as follows: S101. Collect environmental data for all types of geology, identify the types of environmental data that exist in all types of geology, and use expert knowledge to review and screen the types of environmental data that exist in all types of geology to obtain the common characteristics of all types of geology. S102. During real-time geological surveys, sensors are used to collect general characteristic data of the real-time geological surveys. Characteristic values ​​are obtained by calculating and analyzing the general characteristic data using the following formula: ; In the formula, e i f represents the calculated i-th eigenvalue. i Let represent the i-th feature extraction function, which includes, but is not limited to, the mean slope and water depth; Sen represents the general feature data; for each type of general feature data, feature values ​​are calculated and standardized; S103. Construct an environmental feature matrix E = [e1, e2, e3, ..., e] using all calculated eigenvalues. n ] T e1, e2, e3, ..., e n This represents the 1st, 2nd, 3rd, ..., nth eigenvalues ​​calculated, where n is a positive integer.

[0026] By screening the common environmental data of all geological types (such as soil moisture, rock stress, geomagnetic intensity, etc.), the data fragmentation caused by the difference in geological types is avoided, and a unified foundation is laid for subsequent cross-type analysis. The environmental feature matrix is constructed, which facilitates the rapid processing of massive real-time data by computer and provides structured data support for subsequent real-time analysis.

[0027] S200, according to professional knowledge, a knowledge base of the common feature range in each type of geology is predefined; the similarity of the real-time environmental feature matrix and the geological data of each type in the knowledge base is calculated, and the real-time survey geological type is selected and judged; The specific steps for selecting and judging the real-time survey geological type are as follows: S201, according to professional knowledge, the common feature range in each type of geology is predefined, a geological database containing all types is constructed, a type label g is generated for each type of geology in the geological database, the average value and standard deviation of different common data are calculated for each type of geology in the geological database, the similarity of the real-time environmental feature matrix and each type of geology in the geological database is calculated, and the formula is: ; In the formula, Sim k represents the similarity of the real-time environmental feature matrix and the kth type of geology in the geological database, e i represents the ith feature value in the real-time environmental feature matrix, μ ki represents the average value of the ith common feature in the kth type of geology, sd ki represents the standard deviation of the ith common feature in the kth type of geology; S202, a similarity screening mechanism is constructed, and the formula is: ; In the formula, g t represents the type label of the real-time survey geology, and the formula represents the meaning that from all types of geology k, the type label corresponding to k is selected as the type label of the real-time survey geology g k which makes Sim t ; The similarity screening mechanism is used to judge and screen the similarity of the real-time environmental feature matrix and all types of geology in the geological database, and the type label of the real-time survey geology is obtained.

[0028] By calculating the similarity of the real-time data and the knowledge base, the complex geological type can be automatically recognized, and the subjective error of manual interpretation is avoided. The geological type is determined in advance, which provides a "type label" for subsequent data fusion and abnormal judgment, and solves the defects of "one-size-fits-all" analysis of general systems.

[0029] S300, calculate the quality score and correlation of the real-time general feature data, and calculate the data fusion weight of the real-time survey geological type according to the calculated quality score and correlation; The specific steps of calculating the data fusion weight of the real-time survey geological type according to the calculated quality score and correlation are as follows: S301, for the real-time survey geological type label, use expert knowledge to score each general feature in the real-time survey geological type, and calculate the historical accuracy rate of each general feature, the formula is: In the formula, Lz represents the historical accuracy rate of the general feature, Cr represents the number of correct detections in the history of the general feature, and Cz represents the total number of uses of the general feature in the history; and the average of the expert score and the historical accuracy rate of each general feature is taken as the correlation of the general feature. The signal-to-noise ratio, the theoretical maximum signal-to-noise ratio and the data coverage of each general feature in the historical data source are collected, and the quality score of each general feature is calculated, the formula is: ; In the formula, q represents the quality score of the general feature, SNR represents the signal-to-noise ratio of the general feature in the historical data source, SNR max represents the theoretical maximum signal-to-noise ratio in the historical data source, and Cover represents the coverage of the general feature. S302, calculate the fusion weight according to the correlation and quality score of each general feature, the formula is: ; In the formula, w j represents the fusion weight of the jth general feature, Rel j represents the correlation of the jth general feature, q j represents the quality score of the jth general feature; m represents the total number of general features, and the fusion weight of all general features is calculated repeatedly.

[0030] In combination with the quality score and the correlation, noise data (such as abnormal values caused by sensor failure) is removed, and data reliability is improved. Through dynamic weight calculation, the weight of the anti-interference sensor (such as the optical fiber strain sensor) is automatically increased for special geological environment (such as the high magnetic interference area), the influence of the disturbed data (such as the traditional electromagnetic sensor) is reduced, and the system adaptability is enhanced.

[0031] S400, when surveying geology in real time, use sensors to collect environmental data to generate real-time data streams, and use the calculated data fusion weight to perform weighted fusion on the real-time data streams when surveying geology; The specific steps of using the calculated data fusion weight to perform weighted fusion on the real-time data streams when surveying geology are as follows: S401, in real-time surveying geology, collecting environment data by sensors to generate real-time data stream, extracting real-time general feature data value in the real-time data stream, weighting fusion by fusion weight, formula is: ; In the formula, Rt represents the fusion data after weighting fusion, Sd j represents the real-time data value of the jth general feature.

[0032] Weighting fusion of real-time data stream can weaken sudden interference in real time and ensure data continuity. Fusion of multi-dimensional data such as geology, hydrology and meteorology forms a "three-dimensional geological portrait" and solves the one-sidedness of single data dimension analysis. Through weight adjustment, data real-time is ensured (such as second-level update), and error accumulation caused by rapid collection is avoided.

[0033] S500, according to different types of geological history survey data, calculating the abnormal threshold of different types of geology and standardizing to generate abnormal threshold database; The specific steps of generating abnormal threshold database are: S501, for each type of geology, extracting feature values in the historical environment feature matrix and standardizing, calculating the average value and standard deviation of the historical feature values, calculating the environmental variation coefficient of each type of geology by feature value, formula is: ; In the formula, EnvVar represents the environmental variation coefficient of each type of geology, e i represents the ith feature value in the real-time environment feature matrix, μ i represents the average value of the ith historical feature value, sd i represents the standard deviation of the ith historical feature value; S502, extracting the average value and standard deviation of each type of geology after fusion in history and standardizing, calculating the abnormal threshold of each type of geology, formula is: ; In the formula, τ k represents the abnormal threshold of the kth type of geology, Rμ k represents the average value of the historical fusion data of the kth type of geology, Rsd k represents the standard deviation of the historical fusion data of the kth type of geology, and α represents the sensitivity coefficient; All types of geology are used to construct the abnormal threshold database.

[0034] Based on the history survey data of different types of geology, the abnormal standard more suitable for the actual scene is established to avoid misjudgment caused by general threshold.

[0035] S600, extracting the abnormal threshold of the real-time survey geology type in the abnormal threshold database, and judging the fusion data by using the abnormal threshold to obtain whether the real-time survey geology exists abnormality and early warning.

[0036] The specific steps of obtaining whether the real-time survey geology exists abnormality and early warning are as follows: S601, extracting the abnormal threshold τ of the real-time survey geology type in the abnormal threshold database gt , judging the fusion data of the real-time survey geology by using the abnormal threshold of the real-time survey geology type, when Rt>τ gt , judging that the real-time survey geology exists abnormal risk and early warning; when Rt≤τ gt , judging that the real-time survey geology does not exist abnormal risk.

[0037] Based on the extraction of corresponding threshold according to the real-time geology type (for example, the threshold of karst cave development of karst landform is different from the threshold of landslide of the Loess Plateau), the false negatives caused by "unified standard" of the general system are reduced. Combined with the fusion data and the threshold, multi-level early warning (for example, yellow early warning→red early warning) can be realized, and a time window is provided for emergency decision-making.

[0038] The geological survey data analysis system based on multi-source data fusion includes a data acquisition module, a feature analysis module, a survey type judgment module, a fusion module, an abnormal threshold module and an abnormal judgment module. The data acquisition module is used to collect environmental data of all types of geology, and extract feature values in the historical environmental feature matrix; The feature analysis module is used to extract general features of all types of geology, and calculate and analyze the general feature data to obtain feature values; The survey type judgment module is used to calculate the similarity between the real-time feature value and each type of geology in the geology database, and filter the maximum value to obtain the real-time survey geology type; The fusion module is used to calculate the fusion weight of each general feature, and perform weighted fusion on the general features in the real-time data stream by using the fusion weight; The abnormal threshold module is used to calculate the abnormal threshold by using the average value, standard deviation and environmental variation coefficient of the historical fusion data; The abnormal judgment module is used to judge the fusion data of the real-time survey geology by using the abnormal threshold, and obtain whether the real-time survey geology exists abnormality and early warning.

[0039] The feature analysis module includes a general feature unit and a feature value unit; The general feature unit is used to review and filter the types of environmental data that exist in all types of geology by using expert knowledge, and obtain the general features of all types of geology; The characteristic value unit is used for collecting general characteristic data of real-time survey geology by using sensors, and calculating and analyzing the general characteristic data to obtain characteristic values.

[0040] The fusion module comprises a fusion weight unit and a data fusion unit. The fusion weight unit is used for calculating the correlation degree and quality score of each general characteristic, and calculating the fusion weight of each general characteristic. The data fusion unit is used for collecting environmental data by using sensors to generate a real-time data stream, extracting real-time general characteristic data values in the real-time data stream, and performing weighted fusion by using the fusion weight.

[0041] Embodiment 1: obtaining a knowledge base according to expert knowledge

[0042] Table 1 Embodiment 2: assuming that the general characteristics are magnetic field and slope, and the correlation degrees of each general data for mine geology are 0.9 and 0.2, the quality scores of the magnetic field and slope general characteristic data sources are 0.782 and 0.80. The fusion weight of the magnetic field general characteristic is 0.81, 0.19. It is apparent for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, but rather can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all respects as illustrative and not restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended to embrace all changes falling within the meaning and scope of the equivalent elements of the claims. Any reference signs in the claims should not be considered as limiting the claims involved.

Claims

1. A geological exploration data analysis method based on multi-source data fusion, characterized in that: The method includes the following steps: S100. Screen environmental data from all types of geology, extract all existing environmental data and obtain common features through professional knowledge review, collect common feature data from geology using multi-source sensors, calculate real-time feature values ​​and construct an environmental feature matrix. S200: Generate a knowledge base by predefining the common feature range of each type of geology based on professional knowledge; calculate the similarity between the real-time environmental feature matrix and the geological data of each type in the knowledge base, and select and judge to obtain the real-time geological survey type; S300. Calculate the quality score and relevance of real-time general feature data, and use the calculated quality score and relevance to calculate the data fusion weight of real-time geological survey types respectively; S400. During real-time geological surveys, environmental data is collected using sensors to generate a real-time data stream. The calculated data fusion weights are then used to perform weighted fusion of the real-time data stream during geological surveys. S500: Based on different types of geological historical survey data, calculate and standardize the anomaly thresholds for different types of geology to generate an anomaly threshold database. S600. Extract the anomaly threshold of the real-time geological survey type from the anomaly threshold database, and use the anomaly threshold to judge the fused data to determine whether there are anomalies in the real-time geological survey and issue an early warning.

2. The geological exploration data analysis method based on multi-source data fusion according to claim 1, characterized in that: The specific steps for calculating real-time feature values ​​and constructing the environmental feature matrix in S100 are as follows: S101. Collect environmental data for all types of geology, identify the types of environmental data that exist in all types of geology, and use expert knowledge to review and screen the types of environmental data that exist in all types of geology to obtain the common characteristics of all types of geology. S102. During real-time geological surveys, sensors are used to collect general characteristic data of the real-time geological surveys. Characteristic values ​​are obtained by calculating and analyzing the general characteristic data using the following formula: ； In the formula, e i f represents the calculated i-th eigenvalue. i Let represent the i-th feature extraction function, which includes, but is not limited to, the mean slope and water depth; Sen represents the general feature data; for each type of general feature data, feature values ​​are calculated and standardized; S103. Construct an environmental feature matrix E = [e1, e2, e3, ..., e] using all calculated eigenvalues. n ] T e1, e2, e3, ..., e n This represents the 1st, 2nd, 3rd, ..., nth eigenvalues ​​calculated, where n is a positive integer.

3. The geological exploration data analysis method based on multi-source data fusion according to claim 2, characterized in that: The specific steps for determining the real-time geological type in S200 are as follows: S201. Based on professional knowledge, predefine the common feature range for each type of geology, construct a geological database containing all types, and generate a type label g for each type of geology in the geological database. Within the geological database, calculate the mean and standard deviation of different common data for each type of geology, and calculate the similarity between the real-time environmental feature matrix and each type of geology in the geological database. The formula is: ； In the formula, Sim k E represents the similarity between the real-time environmental feature matrix and the k-th type of geology in the geological database. i μ represents the i-th eigenvalue in the real-time environment feature matrix. ki sd represents the average value of the i-th general characteristic in the k-th geological type. ki This represents the standard deviation of the i-th general characteristic in the k-th geological type; S202. Construct a similarity filtering mechanism, the formula of which is: ； In the formula, g t This represents the type label for real-time geological surveys. The formula means selecting from all geological types k that Sim makes available. k The type label corresponding to the maximum value k is used as the real-time geological type label g. t ; A similarity filtering mechanism is used to judge and filter the similarity between the real-time environmental feature matrix and all types of geology in the geological database, so as to obtain the type label of the real-time surveyed geology.

4. The geological exploration data analysis method based on multi-source data fusion according to claim 3, characterized in that: The specific steps in S300 for calculating the data fusion weights of real-time geological survey types using the calculated quality score and correlation are as follows: S301. For real-time geological type labels, expert knowledge is used to score each common feature in the real-time geological type, and the historical accuracy of each common feature is calculated using the following formula: In the formula, Lz represents the historical accuracy of the general feature, Cr represents the number of correct detections of the general feature in history, and Cz represents the total number of times the general feature has been used in history; the average of the expert score and historical accuracy of each general feature is calculated as the relevance of the general feature; Collect the signal-to-noise ratio, theoretical maximum signal-to-noise ratio, and data coverage of each general feature from historical data sources, and calculate the quality score for each general feature using the following formula: ； In the formula, q represents the quality score of the general feature, and SNR represents the signal-to-noise ratio of the general feature in the historical data source. max This represents the theoretical maximum signal-to-noise ratio in historical data sources, and Cover represents the coverage of general features; S302. Calculate the fusion weight using the relevance and quality score of each general feature, using the following formula: ； In the formula, w j Represents the fusion weight of the j-th general feature, Rel j q represents the relevance of the j-th general feature. j represents the quality score of the j-th general feature; m represents the total number of general features. The fusion weights of all general features are calculated by repeating the calculation.

5. The geological exploration data analysis method based on multi-source data fusion according to claim 4, characterized in that: The specific steps in S400 for weighted fusion of real-time data streams during geological surveys using calculated data fusion weights are as follows: S401. During real-time geological surveys, environmental data is collected using sensors to generate a real-time data stream. Real-time common feature data values ​​are extracted from the real-time data stream, and weighted fusion is performed using fusion weights. The formula is as follows: ； In the formula, Rt represents the weighted fused data, and Sd j This represents the real-time data value of the j-th general feature.

6. The geological exploration data analysis method based on multi-source data fusion according to claim 5, characterized in that: The specific steps for generating the anomaly threshold database in S500 are as follows: S501. For each geological type, extract and standardize the eigenvalues ​​from the historical environmental feature matrix, calculate the mean and standard deviation of the historical eigenvalues, and use the eigenvalues ​​to calculate the environmental variation coefficient for each geological type. The formula is as follows: ； In the formula, EnvVar represents the environmental variability coefficient for each type of geology, e i μ represents the i-th eigenvalue in the real-time environment feature matrix. i sd represents the average value of the i-th historical feature. i This represents the standard deviation of the i-th historical feature value; S502. Extract the average and standard deviation of the common data for each type of geological formation from the historical data, fuse them, and standardize them. Calculate the anomaly threshold for each type of geological formation using the following formula: ； In the formula, τ k Rμ represents the anomaly threshold for the k-th type of geology. k Rsd represents the average historical fusion data of the k-th geological type. k denoted by , where represents the standard deviation of the historical fusion data for the k-th geological type, and α represents the sensitivity coefficient; An anomaly threshold database is constructed using anomaly thresholds for all types of geology.

7. The geological exploration data analysis method based on multi-source data fusion according to claim 6, characterized in that: The specific steps in S600 for obtaining real-time geological anomalies and issuing early warnings are as follows: S601. Extract the anomaly threshold τ for real-time geological survey types from the anomaly threshold database. gt The anomaly threshold of real-time geological survey types is used to judge the fused data of real-time geological surveys. When Rt>τ gt When an anomaly risk is detected in the real-time geological survey, an early warning is issued; when Rt≤τ gt At that time, it was determined that there were no abnormal risks in the real-time geological survey.

8. A geological exploration data analysis system based on multi-source data fusion, characterized in that: The geological exploration data analysis system includes a data acquisition module, a feature analysis module, an exploration type determination module, a fusion module, an anomaly threshold module, and an anomaly judgment module. The data acquisition module is used to collect environmental data of all types of geology and extract feature values ​​from the historical environmental feature matrix. The feature analysis module is used to extract common features of all types of geology and to calculate and analyze the common feature data to obtain feature values. The survey type determination module is used to calculate the similarity between real-time feature values ​​and each type of geology in the geological database, and filter the maximum value to obtain the real-time survey geological type. The fusion module is used to calculate the fusion weight of each general feature, and to perform weighted fusion of the general features in the real-time data stream using the fusion weight; The anomaly threshold module is used to calculate the anomaly threshold using the average value, standard deviation and environmental coefficient of variation of historical fused data; The anomaly detection module is used to judge the fused data of real-time geological surveys using anomaly thresholds, to determine whether there are anomalies in the real-time geological surveys and to issue an early warning.

9. The geological exploration data analysis system based on multi-source data fusion according to claim 8, characterized in that: The feature analysis module includes general feature units and feature value units; The general feature unit is used to review and filter the types of environmental data that exist in all types of geology using expert knowledge, so as to obtain the general features of all types of geology. The feature value unit is used to collect general feature data of real-time geological surveys using sensors, and to calculate and analyze the general feature data to obtain feature values.

10. The geological exploration data analysis system based on multi-source data fusion according to claim 8, characterized in that: The fusion module includes a fusion weighting unit and a data fusion unit; The fusion weight unit is used to calculate the relevance and quality score of each general feature, and to calculate the fusion weight of each general feature; The data fusion unit is used to generate a real-time data stream by collecting environmental data from sensors, extract real-time general feature data values ​​from the real-time data stream, and perform weighted fusion using fusion weights.

Images