Geological exploration data real-time analysis method based on edge calculation

Abstract

The invention discloses a geological exploration data real-time analysis method based on edge calculation, and relates to the technical field of computer data processing, and the method comprises the steps: collecting geological exploration data in real time through a plurality of geological exploration devices, adding a time identifier, a spatial position identifier and a data source identifier, and generating a geological exploration data flow; performing real-time preprocessing and feature extraction operation on the geological exploration data stream in the edge computing node to generate a geological exploration feature data stream; according to the distribution change judgment result, adaptive adjustment is carried out on the analysis parameters of the geological exploration data, and the updated analysis parameters are obtained; based on the updated analysis parameters, performing real-time analysis processing on the geological exploration data partition to generate a geological exploration data analysis result; by performing space-time partitioning and index structure construction on the geological exploration data stream in the edge computing node, an efficient and low-delay edge intelligent analysis solution is provided for geological exploration.

Description

Technical Field

[0001] This invention relates to the field of computer data processing technology, and in particular to a real-time analysis method for geological exploration data based on edge computing. Background Technology

[0002] Against the backdrop of geological exploration technology continuously evolving towards intelligence and real-time capabilities, edge computing, as a distributed architecture that deploys computing, storage, and network resources close to the data source, is gradually becoming a key enabling technology to support efficient data processing in complex field environments. With the widespread application of various types of geological sensors (such as seismic detectors, electromagnetic detectors, gravimeters, etc.), geological exploration systems can continuously collect raw data covering multiple dimensions such as physical fields, chemical composition, and structural characteristics with high spatiotemporal resolution. To cope with the challenges of transmission and processing brought about by massive, high-frequency, and heterogeneous data, the industry generally adopts edge computing nodes to perform localized preprocessing of raw data, such as noise suppression, format standardization, or preliminary feature extraction, thereby reducing dependence on the central cloud platform, reducing communication latency, and improving the speed of on-site decision-making response. Existing edge computing-based geological exploration data processing methods still have significant limitations in their analysis mechanisms. Core analysis parameters are usually fixed during the deployment phase, lacking the ability to perceive and respond to dynamic changes in data flow distribution during the exploration process. When the geological environment undergoes abrupt changes due to factors such as tectonic activity, groundwater migration, or human disturbance, the static parameter configuration is difficult to adapt to the new data statistical characteristics, which can easily lead to feature distortion, anomaly underreporting, or misjudgment, thereby weakening the accuracy and reliability of real-time analysis results. Summary of the Invention

[0003] In view of the aforementioned existing problems, the present invention is proposed.

[0004] Therefore, this invention provides a real-time analysis method for geological exploration data based on edge computing to solve the problem of decreased analysis accuracy caused by static configuration of analysis parameters.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a real-time analysis method for geological exploration data based on edge computing. The method includes: acquiring geological exploration data in real time using various geological exploration devices, and adding time stamps, spatial location stamps, and data source stamps to generate a geological exploration data stream; performing real-time preprocessing and feature extraction operations on the geological exploration data stream in edge computing nodes to generate a geological exploration feature data stream; performing spatiotemporal partitioning of the geological exploration feature data stream according to time range and spatial region, and constructing a geological exploration data index structure for each spatiotemporal partition to obtain geological exploration data partitions; continuously monitoring the geological exploration data partitions in the edge computing nodes and comparing and analyzing the distribution status of the geological exploration feature data streams to generate distribution change judgment results; adaptively adjusting the analysis parameters of the geological exploration data based on the distribution change judgment results to obtain updated analysis parameters; and performing real-time analysis processing on the geological exploration data partitions based on the updated analysis parameters to generate geological exploration data analysis results.

[0006] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for generating the geological exploration data stream are as follows: Geological exploration equipment is deployed at the geological exploration site, and the equipment is controlled to collect geological exploration data in real time. During the collection process, the corresponding collection time is automatically recorded for each piece of geological exploration data as a time identifier, and the corresponding spatial location identifier is matched for the geological exploration data according to the location of the geological exploration equipment. Based on the type and number of the geological exploration equipment used to collect the geological exploration data, a geological exploration data source identifier is added. The geological exploration data with time identifier, spatial location identifier and data source identifier are then packaged and arranged into a continuous data sequence according to the collection order to generate a geological exploration data stream.

[0007] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for generating the geological exploration feature data stream are as follows: The edge computing node receives the geological exploration data stream, and performs sequential sorting and integrity verification on the geological exploration data stream based on the time identifier, spatial location identifier and data source identifier carried in the geological exploration data stream to form a complete geological exploration data stream. Noise suppression, outlier identification and removal, missing data compensation, and data dimension unification are performed on the complete geological exploration data stream to obtain preprocessed geological exploration data. Geological feature parameters are extracted from preprocessed geological exploration data using a statistical distribution feature extraction method, and then associated and encapsulated with corresponding time markers, spatial location markers, and data source markers to form a geological exploration feature data stream.

[0008] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for performing spatiotemporal partitioning of the geological exploration feature data stream according to time range and spatial region are as follows. Based on the corresponding time identifier in the geological exploration feature data stream, and according to the time range division rules, the geological exploration feature data stream is grouped according to continuous time intervals, and the geological exploration feature data in the same time interval is divided into the same time partition. Based on the spatial location identifiers recorded in the geological exploration feature data stream, the geological exploration feature data within the same time partition are mapped to the corresponding spatial regions according to the spatial region division rules, thus forming the corresponding spatial region partitions. Geological exploration feature data from the same time zone and the same spatial region are grouped into spatiotemporal zones, and a unique zone identifier is assigned to each spatiotemporal zone.

[0009] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for obtaining the geological exploration data partitions are as follows: In the edge computing node, for each spatiotemporal partition identifier, the geological exploration feature data associated with the spatiotemporal partition identifier is read, and the time identifier, spatial location identifier, data source identifier and feature parameters in the geological exploration feature data are organized to form an index field set; A geological exploration data index structure is constructed based on the set of index fields for the corresponding spatiotemporal partitions. Geological exploration feature data belonging to the same spatiotemporal partition are associated and encapsulated with the corresponding geological exploration data index structure to form a geological exploration data partition.

[0010] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for generating the distribution change determination result are as follows: In the edge computing node, a corresponding continuous monitoring task is established for the geological exploration data partition, forming a partition monitoring feature data stream. The partition monitoring feature data stream is then subjected to sliding window caching and time alignment processing to obtain a geological exploration feature data set and calculate the current distribution description vector. Historical distribution description vectors with the same caliber as the current distribution description vector are extracted from the geological exploration data partitions and aligned in time and space to obtain comparative distribution description vectors; The current distribution description vector and the comparison distribution description vector are compared and analyzed to generate a set of distribution difference indicators. Combined with the change judgment rules, the change of distribution state is judged and a change type label is generated. The change confidence information is determined based on the change magnitude of the distribution difference index set. The change type label is then associated and encapsulated with the change confidence information to generate the distribution change judgment result.

[0011] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the change determination rule is formed by statistically modeling the distribution changes of historical geological exploration data under different geological environments and working conditions, and then combining it with offline learning from field surveys.

[0012] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the specific steps for obtaining the updated analysis parameters are as follows: The change type label, change confidence information, and spatiotemporal partition identifier corresponding to the change type label are parsed from the distribution change judgment results. The change type label, change confidence information, and spatiotemporal partition identifier are then combined to form parameter adjustment trigger information. Based on parameter adjustment trigger information, the analysis parameter mapping table is called in the edge computing node to determine the adjustable analysis parameter items; Based on the change confidence information, the parameter adjustment magnitude coefficient of the adjustable analysis parameter is calculated, and then weighted and updated with the adjustable analysis parameter to obtain the candidate updated analysis parameter; Constraint verification is performed on the candidate updated analysis parameters to obtain the verified updated analysis parameters. The verified updated analysis parameters are written into the parameter configuration instance corresponding to the spatiotemporal partition identifier and a version number is generated. At the same time, the version number is associated with the spatiotemporal partition identifier to form the updated analysis parameters.

[0013] As a preferred embodiment of the real-time analysis method for geological exploration data based on edge computing described in this invention, the analysis parameter mapping table is constructed by performing offline statistical analysis on the analysis effect of historical geological exploration data under different geological environments, equipment operating conditions and distribution change types, and based on the sensitivity assessment of the influence of analysis parameters on the analysis results.

[0014] As a preferred embodiment of the real-time geological exploration data analysis method based on edge computing described in this invention, the specific steps for generating the geological exploration data analysis results are as follows: The updated analysis parameters are loaded as the effective parameter configuration. Based on the effective parameter configuration, the corresponding geological exploration feature data streams are extracted from the geological exploration data partition in chronological order to form real-time analysis data. Based on the effective parameter configuration, real-time analysis and calculation are performed on the real-time analysis data to generate intermediate analysis results. Based on the time window parameters and spatial correlation parameters in the effective parameter configuration, the intermediate analysis results are discriminated and integrated to form valid analysis conclusions. The valid analysis conclusions are associated and encapsulated with the corresponding time markers, spatial location markers, and data source markers to generate geological exploration data analysis results.

[0015] The beneficial effects of this invention are as follows: By leveraging the localized processing advantages of edge computing, and through spatiotemporal partitioning and indexing of geological exploration data streams within edge computing nodes, efficient organization and rapid retrieval of multi-source heterogeneous data are achieved, improving the timeliness and accuracy of local geological dynamic responses in complex field environments; relying on the real-time perception and closed-loop feedback capabilities of edge computing, the distribution status of geological features is continuously monitored, and analysis parameters are adaptively adjusted based on the distribution change judgment results, providing an efficient, reliable, and low-latency edge intelligent analysis solution for geological exploration while ensuring analysis accuracy. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of a real-time analysis method for geological exploration data based on edge computing.

[0018] Figure 2 A flowchart for generating geological exploration data streams.

[0019] Figure 3 A flowchart for generating geological exploration feature data streams.

[0020] Figure 4 The flowchart for generating spatiotemporal partitioning and data partitioning. Detailed Implementation

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0024] Reference Figures 1-4 This is one embodiment of the present invention, which provides a real-time analysis method for geological exploration data based on edge computing, including the following steps: S1. Geological exploration data is collected in real time using various geological exploration equipment, and time, spatial location and data source identifiers are added to generate a geological exploration data stream.

[0025] S1.1. Deploy geological exploration equipment at the geological exploration site, control the geological exploration equipment to collect geological exploration data in real time, and automatically record the corresponding collection time for each piece of geological exploration data as a time identifier during the collection process, and match the corresponding spatial location identifier for the geological exploration data according to the location of the geological exploration equipment.

[0026] Specifically, after the geological exploration equipment is deployed at the geological exploration site, the equipment is controlled to collect geological exploration data in real time. This data includes observational data reflecting the characteristics of underground physical fields (such as seismic, electromagnetic, gravity, and magnetic data), parameter data reflecting geochemical and environmental conditions, structural and coordinate data describing the relationship between geological structures and spatial locations, imagery and topographic data obtained from remote sensing or surface observation, and equipment operation and location information related to the acquisition process. While collecting geological exploration data, the equipment records the acquisition time for each data point and writes the acquisition time into a time identifier field. The time identifier is used to express the temporal order and time range of the geological exploration data. The location of the geological exploration equipment is used to determine its spatial coordinates or spatial location description information, and this spatial coordinates or spatial location description information is written into a spatial location identifier field. The spatial location identifier is used to express the spatial location where the geological exploration data was generated.

[0027] S1.2. Based on the type and number of the geological exploration equipment used to collect the geological exploration data, add a geological exploration data source identifier. Encapsulate the geological exploration data with the added time identifier, spatial location identifier, and data source identifier, and arrange them into a continuous data sequence according to the collection order to generate a geological exploration data stream.

[0028] Specifically, after deploying geological exploration equipment at the geological exploration site, the equipment is controlled to collect geological exploration data in real time. During the collection process, the collection time of each piece of geological exploration data is automatically recorded as a time identifier. At the same time, a spatial location identifier is matched to the geological exploration data based on the location of the geological exploration equipment. After obtaining the time identifier and spatial location identifier, a geological exploration data source identifier is added based on the geological exploration equipment type and equipment number used to collect the geological exploration data, forming geological exploration data that simultaneously contains a time identifier, a spatial location identifier, and a geological exploration data source identifier. The geological exploration data containing the time identifier, spatial location identifier, and geological exploration data source identifier is then encapsulated. Each encapsulated piece of geological exploration data is arranged according to the collection order, and the encapsulated geological exploration data arranged in the collection order are connected into a continuous data sequence. The continuous data sequence generates a geological exploration data stream.

[0029] S2. Perform real-time preprocessing and feature extraction operations on the geological exploration data stream in the edge computing node to generate a geological exploration feature data stream.

[0030] S2.1. Receive geological exploration data stream in the edge computing node, and perform sequential sorting and integrity verification on the geological exploration data stream based on the time identifier, spatial location identifier and data source identifier carried in the geological exploration data stream to form a complete geological exploration data stream.

[0031] Specifically, after receiving the geological exploration data stream in the edge computing node, the system performs sequence sorting and integrity verification based on the time identifier, spatial location identifier, and geological exploration data source identifier carried by each encapsulated record in the geological exploration data stream. Sequence sorting aligns the geological exploration data stream according to the time identifier, and encapsulated records with the same or similar time identifiers are further grouped and arranged according to the geological exploration data source identifier. Integrity verification checks the integrity of the time identifier, spatial location identifier, and geological exploration data source identifier fields for each encapsulated record, while also checking the continuity of the geological exploration data stream, identifying missing records, duplicate records, and out-of-order records, and performing rearrangement for out-of-order records, deduplication for duplicate records, and missing record marking for missing records. The system outputs a complete geological exploration data stream with continuous time identifiers, locatable spatial location identifiers, and traceable geological exploration data source identifiers.

[0032] S2.2. Perform noise suppression, outlier identification and removal, missing data compensation, and data dimension unification on the complete geological exploration data stream to obtain preprocessed geological exploration data.

[0033] Specifically, the complete geological exploration data stream undergoes the following processing steps in sequence: noise suppression, outlier identification and removal, missing data compensation, and data dimension unification. Noise suppression smooths the data stream based on its temporal continuity, reducing the impact of fluctuations in data acquisition from geological exploration equipment. Outlier identification and removal identifies and removes outlier points based on the statistical distribution characteristics of the data stream (obtained from the original sampled data within a certain time window or spatial region) and outlier sampling points. Missing data compensation locates missing sampling positions according to time and spatial identifiers, and fills in the missing positions using interpolation of adjacent sampling points or trend extrapolation based on continuous sampling segments. Data dimension unification performs unit conversion and dimension standardization on the geological exploration data stream according to the unit of measurement and dimensional expression method corresponding to the geological exploration data source identifier, making the data stream comparable under a unified dimension, resulting in preprocessed geological exploration data.

[0034] S2.3. Geological feature parameters are extracted from the preprocessed geological exploration data using a statistical distribution feature extraction method, and then associated and encapsulated with the corresponding time identifier, spatial location identifier, and data source identifier to form a geological exploration feature data stream.

[0035] Specifically, each piece of preprocessed geological exploration data is read from the preprocessed geological exploration data in chronological order of acquisition. Feature parameters are extracted based on differences in numerical distribution, rate of change, fluctuation amplitude, spectral energy distribution, and spatial gradient. These feature parameters are used to characterize the geological structural state, geological anomaly characteristics, and geological change trends. For each set of feature parameters, the time identifier, spatial location identifier, and data source identifier carried in the preprocessed geological exploration data are matched synchronously. The feature parameters are then associated and encapsulated with the time identifier, spatial location identifier, and data source identifier. The encapsulation results are arranged continuously in the acquisition order to form a geological exploration feature data stream.

[0036] S3. Perform spatiotemporal partitioning of the geological exploration feature data stream according to time range and spatial region, and construct a geological exploration data index structure for each spatiotemporal partition to obtain the geological exploration data partition.

[0037] S3.1. Based on the corresponding time identifier in the geological exploration feature data stream, and according to the time range division rules, the geological exploration feature data stream is grouped according to continuous time intervals, and the geological exploration feature data in the same time interval is divided into the same time partition.

[0038] Specifically, the process involves reading the time stamp corresponding to each geological exploration feature data entry within the geological exploration feature data stream and arranging the data stream according to the order of the time stamps. The boundaries of continuous time intervals are determined based on time range division rules, which are then used as the grouping criterion. These rules combine start and end times with interval length, for example, using 0 seconds as the start time and 60 seconds as the interval length. Within the arranged geological exploration feature data stream, each time stamp falls within a specific continuous time interval. Geological exploration feature data with time stamps between 0 and 60 seconds are assigned to the first time partition, and those between 60 and 120 seconds are assigned to the second time partition. After this process, multiple time partitions are formed, distinguished by continuous time interval boundaries. Each time partition contains geological exploration feature data within the coverage area of ​​the continuous time interval.

[0039] It should be noted that the time range division rules are obtained through offline statistical analysis of the acquisition time characteristics of historical geological exploration data. The statistical content includes the acquisition interval distribution of time markers in the geological exploration feature data stream, the length of continuous sampling segments, and the time scale of distribution changes. Based on the statistical analysis, combined with the requirements of geological exploration tasks for real-time analysis response speed and stability, the length of the time interval that can cover the complete sampling features and is sensitive to distribution changes is selected, and the start and end methods of the time range and the boundaries of continuous time intervals are determined accordingly, forming the time range division rules for grouping and processing geological exploration feature data streams.

[0040] S3.2. Based on the spatial location identifiers recorded in the geological exploration feature data stream, the geological exploration feature data within the same time partition are mapped to the corresponding spatial regions according to the spatial region division rules to form the corresponding spatial region partitions.

[0041] Specifically, within each time partition, the geological exploration feature data stream records contained in the time partition are read, and the spatial location identifiers corresponding to each geological exploration feature data stream record are extracted one by one. The spatial location identifiers are mapped and matched to spatial regions according to the spatial region division rules, and the mapping results are written into spatial region identifiers. The spatial region division rules are used to limit the boundaries of spatial regions and the generation method of spatial region identifiers. For example, the spatial region division rules use latitude and longitude grid division, with an example value of 10×10 meter grid. After the spatial location identifier falls into the corresponding grid, the grid number is associated as the spatial region identifier. After the mapping and matching are completed, the geological exploration feature data stream records in the same time partition are collected according to the spatial region identifiers, and the collection results form multiple spatial region partitions.

[0042] It should be noted that the spatial region division rules are obtained through offline statistical analysis of the spatial location distribution of historical geological exploration data. The statistical content includes the distribution density of coordinate points corresponding to spatial location identifiers, the range of spatial correlation, and the scale of change of geological structural features in spatial dimensions. Based on the statistical analysis results, combined with the boundaries of geological structural units in the geological exploration area, the density of geological exploration equipment deployment, and the requirements for analysis accuracy, the granularity of spatial region division and the expression method of regional boundaries are determined, thereby forming spatial region division rules for spatial mapping processing of geological exploration feature data.

[0043] S3.3. Collect geological exploration feature data from the same time zone and the same spatial region into spatiotemporal zones, and assign a unique zone identifier to each spatiotemporal zone.

[0044] Specifically, geological exploration feature data within the same time range and mapped to the same spatial region are aggregated and processed. The aggregation process uses the time region to which the time identifier belongs and the spatial region region to which the spatial location identifier belongs as aggregation conditions. The aggregation result forms a spatiotemporal partition. For each spatiotemporal partition, a unique partition identifier is generated by combining and encoding the time partition identifier and the spatial region partition identifier.

[0045] S3.4. In the edge computing node, for each spatiotemporal partition identifier, read the geological exploration feature data associated with the spatiotemporal partition identifier, and organize the time identifier, spatial location identifier, data source identifier and feature parameters in the geological exploration feature data to form an index field set.

[0046] Specifically, in the edge computing node, a read operation is performed for each spatiotemporal partition identifier. Based on the spatiotemporal partition identifier, the set of geological exploration feature data records associated with the spatiotemporal partition identifier is located, and the time identifier, spatial location identifier, data source identifier, and feature parameters contained in the set of geological exploration feature data records are extracted one by one. The time identifier is sorted into time order, the spatial location identifier is sorted into spatial region, the data source identifier is sorted into equipment source, and the feature parameters are sorted into field names and field values, and the results are summarized to form an index field set.

[0047] S3.5. Construct a geological exploration data index structure for the corresponding spatiotemporal partition based on the set of index fields, and associate and encapsulate the geological exploration feature data belonging to the same spatiotemporal partition with the corresponding geological exploration data index structure to form a geological exploration data partition.

[0048] Specifically, when constructing a geological exploration data index structure for spatiotemporal partitions based on the index field set, the edge computing nodes generate mapping records between index entries and storage locations according to the field order of time identifier, spatial location identifier, data source identifier, and feature parameters in the index field set, and write the index entries into the geological exploration data index structure. The edge computing nodes associate and encapsulate geological exploration feature data belonging to the same spatiotemporal partition with the geological exploration data index structure. The encapsulation content includes the spatiotemporal partition identifier, the geological exploration data index structure, and the geological exploration feature data set, forming a geological exploration data partition.

[0049] S4. Continuously monitor the geological exploration data partitions in the edge computing nodes, compare and analyze the distribution status of the geological exploration feature data streams, and generate distribution change judgment results.

[0050] S4.1. Establish corresponding continuous monitoring tasks for geological exploration data partitions in edge computing nodes, form partition monitoring feature data streams, perform sliding window caching and time alignment processing on partition monitoring feature data streams, obtain geological exploration feature data sets, and calculate the current distribution description vector.

[0051] Specifically, in edge computing nodes, corresponding continuous monitoring tasks are established for geological exploration data partitions. The continuous monitoring tasks take the geological exploration data partitions as input and output the partition monitoring feature data streams in the order of collection. Sliding window caching is performed on the partition monitoring feature data streams to write geological exploration feature data within a continuous time range into the same cache window. Within the sliding window, the geological exploration feature data are time-aligned according to the time identifier, so that geological exploration feature data from different sources are arranged under a unified time benchmark, forming a geological exploration feature data set. Based on the geological exploration feature data set, the feature parameters are summarized and arranged to obtain the current distribution description vector.

[0052] S4.2. Extract historical distribution description vectors with the same caliber as the current distribution description vector from the geological exploration data partitions, and align them in time and space to obtain comparative distribution description vectors.

[0053] Specifically, historical geological exploration feature data with the same set of feature parameters, time identifier granularity, spatial location identifier granularity, and data source identifier caliber as the current distribution description vector are retrieved from the geological exploration data partition. Based on the time identifier, historical geological exploration feature data falling within the target time range (using the sliding window time interval corresponding to the current distribution description vector as a reference, selecting a historical time interval with consistent time length, time granularity, and statistical caliber from the historical geological exploration feature data) are selected. The boundary of the target time range is matched and aligned with the time boundary corresponding to the sliding window cache of the partition monitoring feature data stream. When there is an interval between the time identifier of the historical geological exploration feature data and the target time boundary, resampling or linear interpolation is used to supplement the feature parameter values ​​at the alignment time. For example, the time step is 10 seconds. Based on the spatial location identifier, the historical geological exploration feature data is mapped to a spatial region partition consistent with the geological exploration data partition, and the feature parameter distribution is summarized according to the spatial region partition to form a comparative distribution description vector.

[0054] S4.3. Perform a distribution state comparison analysis between the current distribution description vector and the comparison distribution description vector, generate a set of distribution difference indicators, and combine them with the change judgment rules to determine the change in distribution state and generate change type labels.

[0055] Specifically, when performing a distribution state comparison analysis between the current distribution description vector and the comparison distribution description vector, the consistency of the vector dimensions, feature parameter order, time identifier caliber, and spatial location identifier caliber of the current and comparison distribution description vectors is checked to obtain directly comparable current and comparison distribution description vectors. A one-way difference calculation is performed on the current and comparison distribution description vectors to obtain one-way difference vectors. Based on these one-way difference vectors, a set of distribution difference indicators is calculated. This set includes vector Euclidean distance, vector cosine similarity change, the mean of the one-way difference absolute value, and the maximum of the one-way difference absolute value. Multiple indicators are uniformly mapped to a single judgment quantity, and each indicator is determined based on its corresponding historical... The historical reference range is normalized to obtain difference values ​​under a unified dimension. The normalized indicators are then weighted and fused to obtain a fused difference score. The fused difference score is used as the basis for determining the distribution state. The set of distribution difference indicators is input into the change determination rule for threshold and interval determination. Based on the change determination rule, the correspondence between the set of distribution difference indicators and the change type label is generated to produce a change type label. When the set of distribution difference indicators is greater than the example value of 0.30, the change type label "mutation" is output. When the set of distribution difference indicators is in the range of the example value of 0.10 to the example value of 0.30, the change type label "gradual change" is output. When the set of distribution difference indicators is less than the example value of 0.10, the change type label "stable" is output.

[0056] It should be noted that the expression for calculating the Euclidean distance between vectors is: ; in, Represents Euclidean distance. This represents the current distribution description vector. Represents a vector describing the historical distribution. The first element of the current distribution description vector represents the... One portion, The first vector representing the historical distribution description vector One portion, Indicates the index of a vector component. Represents the vector dimension.

[0057] The expression for calculating the change in cosine similarity is: ; in, This represents the change in cosine similarity. This represents the dot product of the current distribution description vector and the historical distribution description vector. This represents the L2 norm of the current distribution description vector. This represents the L2 norm of the historical distribution description vector.

[0058] S4.4. Determine the change confidence information based on the change range of the distribution difference index set, associate and encapsulate the change type label with the change confidence information, and generate the distribution change judgment result.

[0059] Specifically, when determining the confidence level information of change based on the change range of the distribution difference index set, the lower limit and upper limit of the reference range corresponding to the distribution difference index set are obtained. The reference range is obtained by statistics from the historical distribution difference index set. The relative position ratio of the change range within the reference range is calculated. Based on the difference between the change range and the lower limit of the reference range, and using the difference between the upper limit and the lower limit of the reference range as the normalized denominator, a relative position ratio between 0 and 1 is obtained. The relative position ratio is used as the basis for the value of the change confidence information. The closer the relative position ratio is to 1, the closer the change range is to the upper limit of the reference range, and the higher the change confidence information. The closer the relative position ratio is to 0, the closer the change range is to the lower limit of the reference range, and the lower the change confidence information. The change type label is associated and encapsulated with the change confidence information to form the distribution change judgment result.

[0060] It should be noted that the change judgment rule is formed by statistically modeling the distribution changes of historical geological exploration data under different geological environments and working conditions, and then combining it with offline learning from field exploration.

[0061] Furthermore, the absolute value of the change range of the distribution difference index set is calculated, and the relative position of the change range is located within the reference range given by the change judgment rule. The reference range is expressed in the form of segmented intervals. For example, the reference range is divided into [0,10), [10,20), [20,30), and ≥30. The change range falling into [0,10) corresponds to low change confidence information, the change range falling into [10,20) corresponds to medium change confidence information, and the change range falling into [20,30) or ≥30 corresponds to high change confidence information. When the change range is close to the upper limit of the interval, the change confidence information is taken from the higher side of the interval level. For example, the value 25 corresponds to high. The change judgment rules are formed based on the distribution change samples of historical geological exploration data under different geological environments and working conditions. Statistical modeling is used to obtain the reference range of the distribution difference index set and the discrimination boundary of the change type label. The sample labeling and consistency verification are completed in combination with the field exploration records. In the offline learning stage, the labeled historical geological exploration data is used to fit and verify the statistical modeling parameters, and the change judgment rules that can be used for distribution status comparison analysis are output, including the mapping relationship between the change type label set, the segmented threshold of the distribution difference index set and the reference range.

[0062] S5. Based on the distribution change determination results, the analysis parameters of the geological exploration data are adaptively adjusted to obtain updated analysis parameters.

[0063] S5.1. Parse the change type label, change confidence information, and spatiotemporal partition identifier corresponding to the change type label from the distribution change determination result, and combine the change type label, change confidence information, and spatiotemporal partition identifier to form parameter adjustment trigger information.

[0064] Specifically, a field parsing operation is performed on the distribution change judgment result. The change type label and change confidence information are extracted from the encapsulated content of the distribution change judgment result, and the spatiotemporal partition identifier associated with the change type label is read. The field parsing process takes the record unit of the distribution change judgment result as the processing granularity. For each record unit, the change type label format is verified, the value range of the change confidence information is verified, and the uniqueness of the spatiotemporal partition identifier is verified. After the verification is passed, the change type label, change confidence information, and spatiotemporal partition identifier are combined and organized according to the association relationship of the same record unit to form parameter adjustment trigger information.

[0065] S5.2. Based on the parameter adjustment trigger information, call the analysis parameter mapping table in the edge computing node to determine the adjustable analysis parameter items.

[0066] Specifically, based on parameter adjustment trigger information, the edge computing node reads the change type label, change confidence information, and spatiotemporal partition identifier from the parameter adjustment trigger information. It then uses the change type label as the first search key, the interval containing the change confidence information as the second search key, and the spatiotemporal partition identifier as the third search key, performing a multi-key matching search in the analysis parameter mapping table. The analysis parameter mapping table records the mapping relationship between the change type label, change confidence interval, and spatiotemporal partition identifier to adjustable analysis parameter items. After a successful search, it outputs a list of adjustable analysis parameter items and retains the parameter item names for subsequent adaptive adjustment of analysis parameters. For example, when the change type label is "mutation," the example value of the change confidence information is 0.85, and the example value of the spatiotemporal partition identifier is 1053, the analysis parameter mapping table returns examples of adjustable analysis parameter items including "noise suppression strength parameter" and "outlier removal threshold parameter."

[0067] S5.3. Calculate the parameter adjustment magnitude coefficient for the adjustable analysis parameter item based on the change confidence information, and perform a weighted update with the adjustable analysis parameter item to obtain the candidate updated analysis parameters.

[0068] Specifically, the change confidence information serves as the basis for determining the parameter adjustment amplitude coefficient. The change confidence information undergoes interval normalization, mapping it to the parameter adjustment amplitude coefficient. This coefficient characterizes the adjustment strength of the adjustable analysis parameter. The closer the change confidence information is to the upper limit of the reference range, the closer the parameter adjustment amplitude coefficient is to the maximum adjustment strength; the closer the change confidence information is to the lower limit of the reference range, the closer the parameter adjustment amplitude coefficient is to the minimum adjustment strength. Using the current value of the adjustable analysis parameter as the fusion benchmark, the parameter adjustment amplitude coefficient is converted into candidate adjustment weights. These candidate adjustment weights are then used to weight and fuse the current values ​​of the adjustable analysis parameter, forming the candidate updated analysis parameters.

[0069] S5.4. Perform constraint verification on the candidate updated analysis parameters to obtain the verified updated analysis parameters. Write the verified updated analysis parameters into the parameter configuration instance corresponding to the spatiotemporal partition identifier and generate a version number. At the same time, establish an association between the version number and the spatiotemporal partition identifier to form the updated analysis parameters.

[0070] Specifically, after obtaining the candidate updated analysis parameters in the edge computing node, constraint verification is performed based on parameter value range constraints, parameter type constraints, parameter consistency constraints, and spatiotemporal partition identifier association constraints. Out-of-bounds values, type mismatch values, mutually exclusive combination values, and missing mandatory values ​​are marked and corrected or rolled back to the most recent valid parameter configuration instance values, forming the verified updated analysis parameters. The edge computing node writes the verified updated analysis parameters into the parameter configuration instance corresponding to the spatiotemporal partition identifier. During the writing process, the writing time identifier, writing source identifier, and writing result identifier are recorded. At the same time, a version number is generated and an association relationship is established between the version number and the spatiotemporal partition identifier. The association relationship is used to retrieve the corresponding version number by the spatiotemporal partition identifier and locate the parameter configuration instance. The content of the parameter configuration instance associated with the version number and the spatiotemporal partition identifier is used as the updated analysis parameters.

[0071] It should be noted that the analysis parameter mapping table is constructed by conducting offline statistical analysis of the analysis effect of historical geological exploration data under different geological environments, equipment operating conditions and distribution change types, and based on the sensitivity assessment of the influence of analysis parameters on the analysis results.

[0072] Furthermore, historical geological exploration data is grouped and archived according to geological environment, equipment operating conditions, and distribution change type. Within each group, the corresponding adjustable analysis parameter configurations and geological exploration data analysis results are compiled. The geological exploration data analysis results are represented by a unified set of analysis effect indicators, such as example values ​​of 95, 0.8, and 120. Statistical comparisons of the analysis effect indicator sets of different adjustable analysis parameter configurations under the same distribution change type are conducted to form statistical results of parameter influence. Sensitivity assessments are performed on adjustable analysis parameters, and a list of highly sensitive parameters and their adjustment priorities are determined based on sensitivity ranking. A correspondence is established between distribution change type, geological environment, equipment operating conditions, and the list of highly sensitive parameters, forming an analysis parameter mapping table containing fields such as "distribution change type, geological environment, equipment operating conditions, adjustable analysis parameter, adjustment priority, and recommended adjustment direction".

[0073] S6. Based on the updated analysis parameters, perform real-time analysis and processing on the geological exploration data partitions to generate geological exploration data analysis results.

[0074] S6.1. Load the updated analysis parameters as the effective parameter configuration, and extract the corresponding geological exploration feature data streams from the geological exploration data partitions in chronological order according to the effective parameter configuration in the edge computing node to form real-time analysis data.

[0075] Specifically, the updated analysis parameters are loaded as the effective parameter configuration. After the edge computing node completes the reading of the effective parameter configuration, it accesses the geological exploration data partition within the spatiotemporal partition identifier range defined by the effective parameter configuration. Based on the adjustable analysis parameter items and parameter value requirements specified in the effective parameter configuration, the edge computing node extracts the geological exploration feature data stream according to the time identifier sequence within the geological exploration data partition. During the extraction process, the correspondence between the time identifier, spatial location identifier, data source identifier, and feature parameters in the geological exploration feature data stream is maintained without interruption, thus forming real-time analysis data.

[0076] S6.2. Based on the effective parameter configuration, perform real-time analysis calculations on the real-time analysis data, generate intermediate analysis results, and perform discrimination and integration processing on the intermediate analysis results based on the time window parameters and spatial correlation parameters in the effective parameter configuration to form valid analysis conclusions.

[0077] Specifically, based on the configured effective parameters, edge computing nodes perform real-time analysis calculations on the real-time analysis data in chronological order of acquisition. The real-time analysis calculation process, according to the adjustable analysis parameters defined in the effective parameter configuration, performs line-by-line calculations and status judgments on the feature parameters in the geological exploration feature data stream, outputting corresponding intermediate analysis results. These intermediate analysis results are data generated during the real-time analysis calculation of the geological exploration feature data according to the effective analysis parameters, but have not undergone time and spatial integration and judgment. The intermediate analysis results maintain the correlation with the time markers, spatial location markers, and data source markers in the real-time analysis data. The edge computing nodes aggregate and organize the intermediate analysis results according to the time window parameters in the effective parameter configuration, and then, based on the effective parameters... The spatial correlation parameter in the data configuration performs correlation judgment and consistency verification on intermediate analysis results within the same or adjacent spatial regions. It jointly judges the results from both time and space dimensions. In the time dimension, intermediate results falling within the same or adjacent time windows are aggregated, and it is checked whether anomaly markers, difference scores, or change trends remain consistent over a continuous period of time. In the spatial dimension, intermediate results belonging to the same or adjacent spatial regions are correlated, and it is checked whether the same state or change type is continuously distributed in space. Consistency verification is performed on the correlated results. After the time and spatial dimension judgments are completed, the intermediate analysis results that meet the time window conditions and spatial correlation conditions are integrated and output to form valid analysis conclusions.

[0078] S6.3. Associate and encapsulate the valid analysis conclusions with the corresponding time markers, spatial location markers, and data source markers to generate geological exploration data analysis results.

[0079] Specifically, the process reads the time window range and spatial correlation range corresponding to the valid analysis conclusions, and extracts the corresponding time identifier, spatial location identifier, and data source identifier from the correlation information retained during the generation of the valid analysis conclusions. The valid analysis conclusions are used as the core content and are associated one-to-one with the time identifier, spatial location identifier, and data source identifier to complete field-level encapsulation. The encapsulated content clearly expresses the formation time, corresponding spatial location, and source geological exploration equipment of the valid analysis conclusions. After encapsulation, a structurally complete geological exploration data analysis result is generated.

[0080] In summary, this invention leverages the localized processing advantages of edge computing by constructing spatiotemporal partitions and index structures for geological exploration data streams within edge computing nodes. This enables efficient organization and rapid retrieval of multi-source heterogeneous data, significantly improving the timeliness and accuracy of local geological dynamic responses in complex field environments. Furthermore, relying on the real-time perception and closed-loop feedback capabilities of edge computing, it continuously monitors the distribution status of geological features and adaptively adjusts analysis parameters based on distribution change judgments. This provides an efficient, reliable, and low-latency edge intelligent analysis solution for geological exploration while ensuring analytical accuracy.

[0081] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A real-time analysis method for geological exploration data based on edge computing, characterized in that: include, Geological exploration data is collected in real time using various geological exploration equipment, and time, spatial location and data source identifiers are added to generate a geological exploration data stream; Real-time preprocessing and feature extraction operations are performed on the geological exploration data stream in the edge computing node to generate a geological exploration feature data stream. The geological exploration feature data stream is divided into temporal and spatial partitions according to time range and spatial region, and a geological exploration data index structure is constructed for each temporal and spatial partition to obtain the geological exploration data partitions. In edge computing nodes, the geological exploration data partitions are continuously monitored, and the distribution status of geological exploration feature data streams is compared and analyzed to generate distribution change judgment results. Based on the distribution change determination results, the analysis parameters of the geological exploration data partition are adaptively adjusted to obtain updated analysis parameters; Based on the updated analysis parameters, real-time analysis and processing are performed on the geological exploration data partitions to generate geological exploration data analysis results.

2. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for generating the geological exploration data stream are as follows: Geological exploration equipment is deployed at the geological exploration site, and the equipment is controlled to collect geological exploration data in real time. During the collection process, the corresponding collection time is automatically recorded for each piece of geological exploration data as a time identifier, and the corresponding spatial location identifier is matched for the geological exploration data according to the location of the geological exploration equipment. Based on the type and number of the geological exploration equipment used to collect the geological exploration data, a data source identifier is added to the geological exploration data. The geological exploration data with the added time identifier, spatial location identifier, and data source identifier are then packaged and arranged into a continuous data sequence according to the collection order to generate a geological exploration data stream.

3. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for generating the geological exploration feature data stream are as follows: The edge computing node receives the geological exploration data stream and, based on the time identifier, spatial location identifier, and data source identifier, sorts and verifies the integrity of the geological exploration data stream to form a complete geological exploration data stream. Noise suppression, outlier identification and removal, missing data compensation, and data dimension unification are performed on the complete geological exploration data stream to obtain preprocessed geological exploration data. Geological feature parameters are extracted from preprocessed geological exploration data using a statistical distribution feature extraction method, and then associated and encapsulated with corresponding time markers, spatial location markers, and data source markers to form a geological exploration feature data stream.

4. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for performing spatiotemporal partitioning of geological exploration feature data streams according to time range and spatial region are as follows. Based on the corresponding time identifier in the geological exploration feature data stream, and according to the time range division rules, the geological exploration feature data stream is grouped according to continuous time intervals, and the geological exploration feature data in the same time interval is divided into the same time partition. Based on the spatial location identifiers recorded in the geological exploration feature data stream, the geological exploration feature data within the same time partition are mapped to the corresponding spatial regions according to the spatial region division rules, thus forming the corresponding spatial region partitions. Geological exploration feature data of the same time zone and corresponding spatial region zone are aggregated into spatiotemporal zones, and a unique zone identifier is assigned to each spatiotemporal zone.

5. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for obtaining the geological exploration data partitioning are as follows: Read the geological exploration feature data associated with the spatiotemporal partition identifier in the edge computing node, and organize the time identifier, spatial location identifier, data source identifier and feature parameters in the geological exploration feature data to form an index field set; A geological exploration data index structure is constructed based on the set of index fields for the corresponding spatiotemporal partitions. Geological exploration feature data belonging to the same spatiotemporal partition are associated and encapsulated with the corresponding geological exploration data index structure to form a geological exploration data partition.

6. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for generating the distribution change determination result are as follows: In edge computing nodes, corresponding continuous monitoring tasks are established for geological exploration data partitions to form a partition monitoring feature data stream. The zonal monitoring feature data stream is buffered by a sliding window and time-aligned to obtain a set of geological exploration feature data, and the current distribution description vector is calculated. Extract historical distribution description vectors with the same caliber as the current distribution description vector from the geological exploration data partitions, and align them in time and space to obtain comparative distribution description vectors; The current distribution description vector and the comparison distribution description vector are compared and analyzed to generate a set of distribution difference indicators. Combined with the change judgment rules, the change of distribution state is judged and a change type label is generated. The change confidence information is determined based on the change magnitude of the distribution difference index set. The change type label is then associated and encapsulated with the change confidence information to generate the distribution change judgment result.

7. The real-time analysis method for geological exploration data based on edge computing as described in claim 6, characterized in that: The change determination rule is formed by statistically modeling the distribution changes of historical geological exploration data under different geological environments and working conditions, and then combining it with offline learning from field exploration.

8. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for obtaining the updated analysis parameters are as follows: The change type label, change confidence information, and spatiotemporal partition identifier corresponding to the change type label are extracted from the distribution change determination results and used to form parameter adjustment trigger information; Based on parameter adjustment trigger information, the analysis parameter mapping table is called in the edge computing node to determine the adjustable analysis parameter items; Based on the change confidence information, the parameter adjustment magnitude coefficient of the adjustable analysis parameter is calculated, and then weighted and updated with the adjustable analysis parameter to obtain the candidate updated analysis parameter; Constraint verification is performed on the candidate updated analysis parameters to obtain the verified updated analysis parameters. The verified updated analysis parameters are written into the parameter configuration instance corresponding to the spatiotemporal partition identifier and a version number is generated. At the same time, the version number is associated with the spatiotemporal partition identifier to form the updated analysis parameters.

9. The real-time analysis method for geological exploration data based on edge computing as described in claim 8, characterized in that: The analytical parameter mapping table is constructed by conducting offline statistical analysis of the analytical effects of historical geological exploration data under different geological environments, equipment operating conditions, and distribution change types, and based on the sensitivity assessment of the influence of analytical parameters on the analytical results.

10. The real-time analysis method for geological exploration data based on edge computing as described in claim 1, characterized in that: The specific steps for generating the geological exploration data analysis results are as follows: The updated analysis parameters are loaded as the effective parameter configuration. Based on the effective parameter configuration, the corresponding geological exploration feature data streams are extracted from the geological exploration data partition in chronological order to form real-time analysis data. Based on the effective parameter configuration, real-time analysis and calculation are performed on the real-time analysis data to generate intermediate analysis results. Based on the time window parameters and spatial correlation parameters in the effective parameter configuration, the intermediate analysis results are discriminated and integrated to form valid analysis conclusions. The valid analysis conclusions are associated and encapsulated with the corresponding time markers, spatial location markers, and data source markers to generate geological exploration data analysis results.

Images