Geological mineral exploration data processing method based on big data acquisition

Abstract

The invention relates to the technical field of geological mineral exploration and big data analysis and processing, in particular to a geological mineral exploration data processing method based on big data acquisition. The method comprises the following steps: synchronously and structurally extracting exploration decision and cognitive hypothesis information during multi-source data acquisition, converting the exploration decision and cognitive hypothesis information into stage behavior semantic vectors, and embedding the stage behavior semantic vectors into data units to form a stage semantic constraint data set with cognitive background perception ability; a credible interval is dynamically generated by calculating stage cognitive consistency indexes, abnormal data weights are softened, and an enhanced stage credible data set is constructed; on the basis, an exploration stage cognitive evolution mechanism is introduced, and an association network between data is dynamically reconstructed, so that an association structure can be adaptively adjusted along with exploration cognition deepening; and finally, a standardized target expression unit capable of directly supporting prospecting deployment and resource configuration is generated by identifying a stable data structure and self-adaptive granularity expression. According to the method, natural conversion from the original data to the decision support information is realized.

Description

Technical Field

[0001] This invention relates to the field of geological and mineral exploration and big data analysis and processing technology, specifically to a method for processing geological and mineral exploration data based on big data acquisition. Background Technology

[0002] With the continuous development of geological and mineral exploration, remote sensing, geophysical exploration, geochemical exploration and engineering logging have accumulated massive amounts of multi-source and multi-scale data, and a geological big data pattern has been initially formed. In recent years, big data analysis, artificial intelligence and other technologies have developed rapidly, providing new technical approaches for mining hidden patterns from massive exploration data and improving mineral exploration prediction capabilities.

[0003] Chinese invention patent application CN120353874A discloses a multi-source data processing system for geographic information big data, relating to the fields of data acquisition and sensor technology. The system includes: a data acquisition module that connects in real-time to remote sensing satellites, drones, IoT sensors, and social media via multi-source heterogeneous interfaces, with adaptive data format parsing and metadata tagging; a distributed storage module that partitions geographic information data based on a spatiotemporal database and object storage architecture, establishing dynamic spatiotemporal indexes and version control; a data fusion module that uses a multi-source data alignment method with dynamic weight allocation to achieve coordinate system transformation, time series calibration, and semantic knowledge graph matching; and an intelligent analysis module and a security management module.

[0004] Current research trends are focused on more deeply integrating and collaboratively analyzing exploration data from different sources and stages, and exploring how to more effectively incorporate the professional knowledge and decision-making processes of exploration personnel into computational models to enhance the contextual awareness and dynamic adaptability of data processing. Against this backdrop, how to construct an intelligent processing method that can adapt to the phased and progressive cognitive characteristics of mineral exploration, and achieve deep integration of multi-source information, dynamic evolution of correlations, and direct condensation of mineral exploration knowledge at the data level has become an important direction for technological development. This aims to promote the evolution of geological exploration from a traditional experience-driven model to a new model driven by data and knowledge. Summary of the Invention

[0005] The purpose of this invention is to address the problems existing in the background technology by proposing a geological and mineral exploration data processing method based on big data collection.

[0006] The technical solution of this invention: a method for processing geological and mineral exploration data based on big data acquisition, comprising the following specific implementation steps: S1. During the multi-source exploration data acquisition process, the decision-making and cognitive hypothesis information of the current exploration stage is extracted and structured, and transformed into stage behavior semantic vectors and embedded into the corresponding data units to form a stage-based semantic constraint dataset with cognitive background perception capabilities. S2. Based on the staged semantic constraint dataset, calculate the stage cognitive consistency index of the data within the stage, dynamically generate a trust interval and calculate the trust weight of each data point to form an enhanced stage trust dataset. S3. Based on the enhanced stage trusted dataset, an exploration stage cognitive evolution mechanism is introduced to dynamically reconstruct the association network between data and perform adaptive sparsification processing on the association network to form an enhanced adaptive association dataset. S4. Based on the enhanced adaptive association dataset, identify stable local structural units, adaptively determine the expression granularity of each structural unit according to the requirements of the mineral exploration task, generate corresponding multi-granularity features, and finally integrate them into a standardized mineral exploration target expression unit.

[0007] Preferably, in step S1, the multi-source exploration data includes: Remote sensing image data, geophysical exploration data, geochemical sampling data, and engineering geological logging data collected simultaneously during the same exploration phase; For each data entry, record its exploration stage identifier, spatial location attribute, and data collection time attribute; The stage behavior semantic vector is generated by extracting and structuring the exploration design description, anomaly delineation basis, causal hypothesis direction and verification priority information for that stage. By embedding stage-specific behavioral semantic vectors into individual data points of the same stage through a behavioral semantic mapping function, extended data units with interpretive constraints are formed and aggregated into stage-specific semantic constraint datasets with clear cognitive boundaries according to the exploration stage.

[0008] Preferably, in step S2, the specific indicators of stage-specific cognitive consistency in calculating data within the stage include: For each data point in the staged semantic constraint dataset, extract its original set of observations and embedded behavioral semantic mapping, and obtain the relevant data set in its spatial or causal neighborhood; The consistency evaluation function is used to comprehensively calculate the consistency relationship between the original observations, behavioral semantic mapping and neighborhood data of the data, and output a numerical index that represents the degree of cognitive matching with the current stage.

[0009] Preferably, in step S2, dynamically generating the confidence interval specifically includes: The mean and standard deviation of the consistency index of stage cognition for all data within the current exploration stage are calculated. Based on the mean and standard deviation, and using the adjustment coefficient to control the expansion or convergence of the interval, the upper and lower bounds of the confidence interval representing the confidence level of the data at this stage are calculated.

[0010] Preferably, in step S2, the process of calculating the credible weight specifically includes: First, calculate the deviation of the stage cognition consistency index of each data point from the mean of the stage consistency index; Then, an exponential decay function is used to calculate the confidence weight of each data point based on the deviation, so that the data with a larger deviation receives a lower weight, thereby achieving a softening effect on abnormal data. Finally, the original observation data, behavioral semantic mapping, stage consistency index, credibility interval and credibility weight are integrated to form an enhanced data unit, and summarized into an enhanced stage credibility dataset.

[0011] Preferably, in step S3, the dynamic reconstruction of the relationship network between data specifically includes: First, based on the credibility weights, spatial coordinates, and attribute vectors containing geophysical, geochemical, and engineering observations of the data units in the enhanced stage credibility dataset, the initial association strength between any two data units is calculated by combining the spatial distance function and the attribute similarity function. Subsequently, information reflecting the cognitive evolution of exploration stages, such as borehole verification results and geological interpretation model revisions, is introduced to generate relevant influencing factors. The initial association strength is dynamically modulated by the stage-based cognitive modulation coefficient to obtain the updated adaptive association strength.

[0012] Preferably, in step S3, the adaptive sparsity processing of the associated network specifically includes: Based on the statistical distribution of all data in the current stage to the updated adaptive association strength, a stage adaptive threshold is dynamically determined to distinguish between strong and weak association edges. For edges identified as weakly correlated, a continuous edge weight decay strategy is adopted, which uses the edge weight decay coefficient to flexibly reduce the weight of weakly correlated edges.

[0013] Preferably, in step S4, identifying stable local structural units specifically includes: Based on the enhanced adaptive association dataset, a sparse association set between each data unit and other data units is extracted to construct a phased weighted association graph; Calculate the structural stability index for each data unit, which is related to the stability of the sum of the weights of its associated edges and the set of its neighborhood. Under the constraint of spatial continuity, multiple data units whose structural stability index exceeds the local structural stability threshold are aggregated to form a local structural unit.

[0014] Preferably, in step S4, adaptively determining the expression granularity of each structural unit specifically includes: For each local structural unit, its granularity fit index is calculated. This index is obtained by combining the internal average confidence weight, internal attribute difference degree, and the mineral exploration task demand intensity index set by the engineering exploration plan of the structural unit through the granularity fit function. Based on the numerical range of the granularity adaptation index, an expression granularity level is adaptively selected for the structural unit from the original observation level, structural feature level, or target decision level.

[0015] Preferably, in step S4, the process of generating standardized mineral exploration target representation units specifically includes: Based on the expression granularity level determined for each local structural unit, the corresponding feature mapping function is used to map the features of each data unit within that structural unit; Based on the comprehensive contribution weight of each data unit, the mapped features are weighted and fused to generate the granular adaptive expression vector of the structural unit. Finally, the granular adaptive expression vector, the spatial coverage information of the structural unit, and the overall credibility are integrated to construct a standardized mineral exploration target expression unit, and summarized to form a stage mineral exploration target expression set.

[0016] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: This invention designs a geological and mineral exploration data processing method based on big data acquisition. First, by structurally embedding stage-based exploration decisions and cognitive assumptions into multi-source data, this method achieves a fundamental shift from passive data recording to proactive interpretability, significantly improving the contextual consistency and interpretability of data in big data analysis. This enables massive heterogeneous data to truly support cognitive-driven geological inferences. Second, by constructing dynamic credibility intervals and continuous weighting functions, it achieves softened processing of anomalous data, effectively suppressing random interference and errors while avoiding the loss of potential mineralization information due to arbitrary removal. This aligns with the gradual progression of exploration cognition from shallow to deep, improving the robustness of information utilization. Third, by introducing a stage-based cognitive evolution mechanism to drive the adaptive reconstruction of the data association network, it enables data to... The spatial, attribute, and causal relationships can be dynamically optimized as the project progresses and understanding is updated, forming a living data structure that evolves in sync with exploration thinking. This provides a stable and flexible analytical foundation for mineral exploration prediction. Furthermore, through the adaptive generation of multi-granularity target expression units, the underlying data is intelligently aggregated into decision-making objects that can directly serve project deployment, opening up a critical path from big data analysis to specific exploration actions. This invention supports a cloud-edge collaborative architecture, enabling key data collection and semantic embedding at the edge, and large-scale correlation analysis and model iteration in the cloud. This ensures the quality of source data while fully leveraging the advantages of cloud computing in big data mining and high-performance analysis, achieving an overall organic unity of intelligent data processing, precise decision support, and adaptive exploration process. Attached Figure Description

[0017] Figure 1 This is a diagram illustrating a method for processing geological and mineral exploration data based on big data acquisition, as proposed in this invention. Detailed Implementation

[0018] Example 1, as Figure 1 As shown, the geological and mineral exploration data processing method based on big data acquisition proposed in this invention includes the following specific implementation steps: S1. Utilizing edge nodes, and taking the exploration stage as the core organizational unit, multi-source exploration data, including remote sensing, geophysical exploration, geochemical exploration, and engineering logging, are collaboratively collected. Exploration decision-making and cognitive hypothesis information is simultaneously introduced during the data generation stage. This transforms data that originally only reflected objective observation results into data units with exploration background awareness capabilities. From the source, a basic data structure that can be effectively utilized as exploration understanding evolves is constructed, providing stable input for subsequent data reliability assessment and correlation evolution processing. The specific implementation process is as follows: S11. By dividing the mineral exploration process into several exploration stages with clear technical objectives, remote sensing, geophysical, geochemical, and engineering data are collected synchronously within each stage, focusing on the stage's tasks. Each data point is then labeled with a clear stage identifier, ensuring a unified cognitive context from the outset and preventing interpretation bias caused by mixing data from different stages. Specifically: During the mineral exploration process, the entire exploration process is divided into several consecutive exploration stages according to the exploration design documents and on-site operation arrangements. Each stage corresponds to clear technical objectives and work priorities, such as the regional prospect investigation stage, the anomaly screening stage, the anomaly verification stage, and the engineering control stage. Within each exploration phase, remote sensing images, geophysical exploration data, geochemical sampling data, and engineering and geological logging data are collected simultaneously, in line with the technical objectives of that phase. All data are required to record the exploration phase identifier at the time of collection, so that they have phase attributes at the data level. To ensure traceability of subsequent processing, a staged data representation format is introduced, representing a single data entry as follows: ; in, This represents the i-th original exploration data record collected during the k-th exploration phase; Represents the set of original observation values ​​corresponding to the i-th data, including but not limited to remote sensing pixel grayscale or spectral values, geophysical measurement response values, geochemical element content data, or attribute descriptions in engineering logging; The spatial location attribute corresponding to the i-th data can be determined by plane coordinates, elevation information, sampling unit number, or borehole location. This indicates the collection time attribute of the i-th data item; This represents the stage identifier information for the k-th exploration stage. This identifier is determined by the exploration implementation plan or field operation plan and is used to bind the data to the stage exploration objectives at the time of its generation. S12. During the implementation of each exploration stage, decision-making information such as exploration design schemes, anomaly screening criteria, and causal hypotheses are extracted and structured simultaneously. This transforms the stage-specific cognition implicit in human experience into behavioral semantic elements that can be perceived by the system, providing clear explanatory premises for subsequent data processing. Specifically: During the implementation of each exploration phase, information such as the exploration design description, anomaly delineation basis, causal hypothesis direction and verification priority are collected simultaneously. The decision elements that appear repeatedly and have a binding effect on data interpretation are extracted and structured (these elements are not simple text descriptions, but are transformed into stage behavioral semantic vectors that can represent the exploration cognitive orientation, and are used to describe the focus of the stage on anomaly type, spatial distribution characteristics and causal patterns). The semantic vector representation of stage behavior is as follows: ; in, This represents the semantic vector of the exploration behavior corresponding to the k-th exploration stage; This represents the j-th type of exploration decision element in the k-th exploration stage, which can correspond to the selection preference of anomaly threshold, the focus of attention on cause type, or the setting of verification priority; m represents the number of decision elements contained in the behavioral semantic vector; S13. Map and embed stage-level behavioral semantic information into individual exploration data, so that each data point retains the original observation values ​​while carrying the exploration cognitive constraints corresponding to its generation. This allows the data to reflect the interpretive differences at different stages in subsequent analysis. Specifically: For each piece of data belonging to the same exploration stage, the behavioral semantic vector corresponding to that stage is introduced into the data structure through a mapping function to form an information unit with interpretive constraints; This mapping does not change the original observations, but rather participates in subsequent data processing as an additional constraint, so that the data can reflect the exploration assumptions at the time of its generation when it is invoked. The embedded data is represented as follows: ; in, Indicates in the original data Extended data units embedded with behavioral semantics; Behavioral semantic mapping function; It should be noted that the behavior semantic mapping function This function transforms decision-making and cognitive information at the exploration stage into interpretive constraints applicable to individual exploration data. Its core function is not to alter the original observation results, but rather to add context and judgment criteria to the data. Based on the degree of attention paid to anomaly types, causal hypotheses, and verification priorities during the exploration stage, the function performs scale compression and weight allocation on behavioral semantics, ensuring that different data exhibit interpretive priorities consistent with their context during subsequent processing. In this way, the data processing results can reflect the evolution of exploration cognition, avoiding misjudgments caused by ignoring differences in stage-specific decisions. S14. After completing the behavioral semantic embedding, the data is uniformly collected and organized according to the exploration stage to form a staged dataset with clear cognitive boundaries. This ensures that subsequent data processing is always carried out within a consistent stage-based cognitive framework, conforming to the engineering process of gradually deepening actual exploration results. Specifically: The data units that have completed behavioral semantic embedding are grouped according to the exploration stage to form a stage dataset with clear cognitive boundaries, which is used for subsequent construction of credible intervals and data association evolution processing. The stage dataset is represented as follows: ; in, represents the semantically constrained dataset formed in the k-th exploration stage; n represents the number of data units contained in the k-th exploration stage.

[0019] S2. Based on the staged semantic constraint dataset from step S1, preliminary consistency calculation and anomaly screening are performed using edge nodes, and global trust interval modeling and weight optimization are conducted using the cloud. Specifically, by constructing a staged cognitive consistency index, cognitive alignment is performed on multi-source data within the same stage, trust intervals are dynamically generated, the weights of anomalous data are softened, and an enhanced staged trust dataset is formed. The specific implementation process is as follows: S21. Perform consistency comparison between each data point and the behavioral semantic mapping and spatial neighborhood data within the stage, calculate the stage cognitive consistency index, reflect the degree of data matching in the current exploration stage, provide a quantitative basis for subsequent reliable interval generation, and simultaneously unify and align multi-source data at the cognitive level, specifically as follows: For each data Extract raw observations and behavioral semantic mapping ; Based on spatial neighborhood information Calculate the consistency relationship with neighboring data for all data units within the same exploration block that are within a distance threshold. Taking into account the degree of deviation between observed values ​​and stage-specific cognition, a consistency index is calculated: ; in, This represents the consistency index of stage cognition for the i-th data in the k-th exploration stage. The higher the value, the better the data matches the current stage cognition and neighboring data. This represents the set of data that are spatially or causally related to the i-th data point; This represents the phase consistency evaluation function; It should be noted that the phase consistency evaluation function It is a calculation method used to measure the degree of matching between each piece of exploration data and the semantics of stage behavior and neighboring data at the current stage. It comprehensively considers the original observation value of the data, the stage cognitive goal and spatial or logical related data. By performing weighted analysis on the direction of data anomalies, distribution trends and causal relationships, it outputs a numerical index to represent the degree of conformity between the data and the stage cognition. The higher the value of the index, the stronger the consistency between the data and the cognition of the current stage. It can be directly used for confidence interval generation and weight allocation. S22. Perform statistical analysis on the overall distribution of consistency indicators within the phase, calculate the mean and standard deviation, dynamically generate confidence intervals, and describe the confidence range of the data through the upper and lower bounds of the intervals to achieve adaptive adjustment of confidence, avoid misjudgments caused by fixed thresholds, and retain the potential for phased anomalies. Specifically: For all consistency metrics within the phase Perform the mean and standard deviation Statistical calculation; Define the confidence interval : ; The lower bound is used to distinguish low-confidence data, and the upper bound is used to mark extremely high-match data. in, This represents the mean of the consistency index in the k-th exploration stage, reflecting the overall cognitive center of that stage; The standard deviation of the consistency index at stage k represents the uncertainty or dispersion of cognition at that stage. and This represents the adjustment coefficient, used to control the expansion or convergence of the confidence interval; This represents the credibility interval of the i-th data point in the k-th stage, characterizing the range of credibility of the data under the stage-specific understanding. S23. For data located at the edge of the confidence interval or slightly below the mainstream interval, a continuous weighting function is used to calculate the confidence weight, allowing it to participate in the analysis with a low weight in subsequent processing. This preserves potentially effective information, reflects the gradual correction pattern of exploration cognition, and achieves soft anomaly processing. Specifically: For each data point, calculate the deviation from the cognitive center of the stage: ; The confidence weights of the data are calculated using an exponential decay function: ; in, This indicates the deviation of the consistency index of the i-th data from the cognitive center of the stage; This represents the confidence weight of the i-th data point in the k-th stage, ranging from (0,1], with a lower weight for larger deviations. S24. Integrate the original observation data, behavioral semantics, spatial neighborhood information, stage consistency indicators, confidence intervals, and weights to form an enhanced stage-based confidence data unit, and output a stage-based confidence dataset. This provides a complete, reliable, and interpretable input foundation for subsequent data association evolution and target prediction. Specifically: The update for each data entry indicates: ; All data units in stage k are aggregated into an enhanced dataset: ; in, Represents an enhanced data unit; This represents the enhanced trusted dataset for stage k, which consists of all the enhanced data units for that stage.

[0020] S3. Based on the enhanced trusted dataset formed in step S2, a cognitive evolution mechanism for the cloud-based exploration stage is introduced to dynamically and adaptively reconstruct the spatial, attribute, and causal relationships between multi-source exploration data. Through trusted weight constraints, cognitive feedback adjustment, and continuous decay of weak associations, the data relationship network can be synchronously adjusted as exploration understanding gradually evolves, thereby constructing a relationship structure that combines stability and flexibility. This provides an interpretable and traceable data foundation for subsequent mineral exploration analysis and target prediction. The specific implementation process is as follows: S31. Based on the trusted weights, spatial locations, and multidimensional attribute information of enhanced data units, an initial data association network is constructed. The association strength between data is quantified by integrating spatial proximity and attribute similarity, enabling stable connections between highly trusted, spatially proximate, and attribute-similar data. Simultaneously, the association contribution of low-trust data is automatically reduced, laying a reasonable initial structure for subsequent dynamic reconstruction. Specifically: The enhanced dataset output from step S2 For each data point: credible weight. Spatial coordinates Attribute vectors This includes geophysical, geochemical, and engineering observations; Define the initial data association strength : ; in, This indicates the initial association strength; a higher value indicates that data i and j are more mutually supportive in terms of space, attributes, and reliability. The spatial distance function can be represented by Euclidean distance, geological spatial topological distance, or depth-weighted distance. This represents an attribute similarity function, such as one calculated using standardized Euclidean distance, cosine similarity, or correlation coefficient. S32. Based on the initial correlation network, cognitive update information from the exploration stage is introduced. New cognitive feedback such as borehole verification and interpretation correction are mapped as correlation adjustment factors to dynamically adjust the correlation strength between data. This allows data relationships to strengthen or weaken as exploration understanding evolves, thereby achieving synchronous evolution of data structure and exploration understanding. Specifically: During the advancement of the k-th exploration phase, as borehole verification results, geological interpretation models are revised, or prospecting hypotheses are updated, incremental cognitive information is generated at each stage. This incremental cognitive information describes the cognitive changes in the current stage relative to the previous stage. This cognitive evolution information is represented in a structured form as follows: ; The new understanding strengthens the causal or spatial connection between the two; The new understanding has no significant impact on the original associations; The new understanding weakens or negates the original related assumptions; Initial data association strength Based on this, stage-based cognitive evolution constraints are introduced to adaptively modulate data association relationships, resulting in updated association strengths: ; It should be noted that in the process of reconstructing associations, the stage-based cognitive evolution does not act in isolation, but rather in conjunction with the data credibility weights formed in step S2. , Collaborative constraints, due to the initial data correlation strength With embedded credible weighting factors, the updated association strength is thus determined after cognitive evolution modulation. It also reflects the combined impact of the credibility of the data itself and the rationality of the stage of cognition; this collaborative mechanism avoids the bias caused by relying solely on expert cognition or solely on data statistics, making the correlation reconstruction results more robust and in line with the actual exploration decision-making logic; in, This represents the set of stage-specific cognitive evolution information for the k-th exploration stage, used to describe the exploration knowledge that has been added or revised in the current stage relative to the previous stage. The factor representing the influence of stage-specific cognitive evolution on the reasonableness of the association between the i-th and j-th data points; This represents the updated adaptive association strength under the constraints of stage-based cognitive evolution; This represents the stage-based cognitive modulation coefficient, used to control the magnitude of the impact of stage-based cognitive evolution on data association reconstruction. S33. After completing the cognitive-driven association reconstruction, the overall association network is adaptively sparsified. Low-strength associations are weakened by setting stage thresholds and a continuous decay mechanism, reducing redundant relationships and highlighting key data connections, while preserving the low-weight participation of potentially anomalous data to avoid excessive removal of valuable information in the early stages. Specifically: In adaptive correlation strength Based on this, statistical analysis is performed on all data association edges in the k-th stage to obtain the overall distribution characteristics of the association strength in this stage; based on this distribution, an adaptive threshold for the stage is constructed. This is used to distinguish between strongly associated edges and weakly associated edges; It should be noted that the stage adaptive threshold Instead of using fixed empirical values, the sparsity strategy is dynamically determined based on the statistical characteristics of the correlation strength within a stage, such as by calculating the stage mean or quantile. This allows the sparsity strategy to be automatically adjusted according to different exploration stages and different data density conditions, avoiding subjective bias caused by human intervention. For the identified weakly related edges, a continuous edge weight decay strategy is introduced to flexibly reduce the weight of weak relationships, specifically expressed as follows: ; in, This represents the adaptive weak association discrimination threshold for the k-th stage; This represents the edge weight decay coefficient, with a value range of [value range missing]. This is used to control the degree of participation of weakly related edges in subsequent analysis; This indicates the correlation strength after sparsification and attenuation processing; It should be noted that due to the original weights of weakly related edges... Data credibility weights have already been incorporated in the preceding steps. , Furthermore, the system incorporates stage-based cognitive evolution constraints, thus eliminating the need to introduce new discrimination conditions during sparsification. Edge weight decay is essentially a structural optimization based on the dual constraints of credibility and cognitive consistency, ensuring that the sparsified network maintains a high degree of consistency in stage-based cognition. By using edge weight decay instead of edge deletion, weak association information is retained in the network with low weights, allowing the association to be quickly activated and strengthened in S32 when subsequent stage cognition changes (e.g., new borehole verification supports the association), without requiring the reconstruction of the network structure. This design significantly improves the continuity and stability of the data association structure during multi-stage evolution, avoiding structural breakage caused by frequent reconstruction. S34. The enhanced data units and their associations after adaptive reconstruction and sparsification are uniformly encapsulated to form a phased enhanced adaptive association dataset. This ensures that each data point simultaneously possesses both credibility constraints and a dynamic association structure, providing direct, complete, and interpretable data input for subsequent multi-scale mineral exploration analysis, target classification, and cognitive evolution modeling. Specifically: In sparsity correlation strength Based on this, the structure of each data point in the k-th exploration stage is expanded, and its original enhanced data unit is merged with the corresponding association relationship to form a new enhanced adaptive data unit, represented as: ; ; To facilitate subsequent analysis and rapid retrieval, the set of association relationships was optimized when constructing the enhanced adaptive data unit. Organize and index the data in an orderly manner; sort the relationships from high to low according to their strength and store them in combination with the data index number, so that the system can directly obtain key or weak relationship information without recalculating the relationships. After completing the construction of the single data unit enhancement structure, all enhanced adaptive data units in the k-th stage are aggregated to form a staged enhanced adaptive association dataset, represented as: ; in, This represents the enhanced adaptive data unit of the i-th data in the k-th stage; Represents the set of sparse associations related to data i, and records the adaptive association strength between it and other data units; This represents the enhanced adaptive association dataset for the k-th exploration stage; n represents the number of data units in the current stage.

[0021] S4. For the enhanced adaptive association dataset output in step S3, the cloud identifies stable data structure units, adaptively determines the expression granularity according to the prospecting task, generates multi-granularity features, and forms target expression units. This achieves a natural transformation of data from "point-structure-decision object," enabling the data to be directly used for prospecting deployment, block selection, and resource allocation, while taking into account reliability, structural integrity, and engineering operability. The specific implementation process is as follows: S41. Calculate the structural stability index of each data unit, identify local structural units by combining spatial continuity constraints, and form a data group that is stable in space, attributes, and cognition. This provides a structured foundation for subsequent granularity adaptation and feature generation, ensuring that the analyzed object is both reliable and traceable. Specifically: For enhanced adaptive association datasets For each data unit, extract its set of associated edges. Construct a phased weighted association graph: ; For each data unit i, calculate its structural stability index: ; Will Joint analysis is performed under spatial continuity constraints to form several local structural units: ; in, This represents the neighborhood set of data unit i, that is, the set of units that are significantly related to i; Indicator representing the local structural stability of data unit i; This represents the threshold for local structural stability. This represents the l-th local structural unit, which consists of multiple relatively stable data units; S42. Based on the internal reliability, attribute complexity, and current mineral exploration task requirements of each structural unit, calculate the granularity adaptation index to adaptively determine the expression granularity of the structural unit, including the original observation level, structural feature level, and target decision level. This achieves dynamic matching between the data expression scale and the requirements of the mineral exploration engineering task, improving analysis efficiency and decision accuracy. For each structural unit Calculate the granularity fit index: ; according to The range of intervals is adaptively selected to determine the expression granularity level. Low granularity: Preserving original or near-original data; Medium granularity: forming structural feature representation; High granularity: generating target-level decision representations; in, Representing structural units The average of the confidence weights of all data units within the data unit; Representing structural units Internal attribute variability or dispersion, i.e., the degree of dispersion of attribute vectors of data units. Calculate the variance or standard deviation; This indicates the intensity of demand for mineral exploration tasks at a particular stage, which is set by the engineering exploration plan or work deployment, such as whether it is the verification drilling stage or the preliminary block screening stage. , and This represents the adjustment coefficient in the granularity adaptation function; Representing structural units The particle size fit index; S43. According to the determined granularity level, adaptive feature mapping and weighted fusion are performed on the data within each structural unit to generate multi-granularity expression vectors. The weights combine data reliability and structural centrality to ensure the complete preservation of information within the structure, while realizing the engineering expression of features at different granularities, providing an operational data foundation for mineral exploration target analysis. Specifically: For structural units Constructing granular adaptive representation: ; in, This represents the overall contribution weight of data unit i within the structural unit; Indicates the granularity fit index Feature mapping function for data unit i; Representing structural units Granularity-adaptive representation vector; It should be noted that the feature mapping function It is used to transform the raw data within a structural unit into a feature representation that can be used for analysis according to its adaptive granularity level. It selects different mapping strategies according to the granularity, such as retaining the original observation information at low granularity, extracting structural features at medium granularity, and generating target decision-level summary information at high granularity. At the same time, it maintains the internal correlation and spatial continuity of the data, realizes the effective compression and expression of multi-granularity information, and provides a compact and interpretable feature representation for subsequent mineral exploration decisions. S44. Integrate multi-granularity features with information such as structural spatial range and overall credibility to construct standardized mineral exploration target expression units, forming a phased target expression set, specifically: Construct standardized mineral exploration target expression units: ; A summary of the mineral exploration target expression set for each stage: And distribute it to the edge nodes; in, Representing structural units The spatial coverage or geographical location range; This represents the overall credibility of a structural unit, summarizing the credibility weights of data units within that unit. And combined with local structural stability indicators calculate; This represents the l-th mineral exploration target representation unit; This represents the complete set of mineral exploration target representation units for stage k.

[0022] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for processing geological and mineral exploration data based on big data acquisition, characterized in that, The specific implementation steps include the following: S1. During the multi-source exploration data acquisition process, the decision-making and cognitive hypothesis information of the current exploration stage is extracted and structured, transformed into stage behavior semantic vectors and embedded into the corresponding data units to form a stage-based semantic constraint dataset with cognitive background perception capabilities. S2. Based on the staged semantic constraint dataset, calculate the stage cognitive consistency index of the data within the stage, dynamically generate a trust interval and calculate the trust weight of each data to form an enhanced stage trust dataset. S3. Based on the enhanced stage trusted dataset, an exploration stage cognitive evolution mechanism is introduced to dynamically reconstruct the association network between data and perform adaptive sparsification processing on the association network to form an enhanced adaptive association dataset. S4. Based on the enhanced adaptive association dataset, identify stable local structural units, adaptively determine the expression granularity of each structural unit according to the requirements of the mineral exploration task, generate corresponding multi-granularity features, and finally integrate them into a standardized mineral exploration target expression unit.

2. The geological and mineral exploration data processing method based on big data acquisition according to claim 1, characterized in that, In step S1, the multi-source exploration data includes: Remote sensing image data, geophysical exploration data, geochemical sampling data, and engineering geological logging data collected simultaneously during the same exploration phase; For each data entry, record its exploration stage identifier, spatial location attribute, and data collection time attribute; The stage behavior semantic vector is generated by extracting and structuring the exploration design description, anomaly delineation basis, causal hypothesis direction and verification priority information for that stage. By embedding stage-specific behavioral semantic vectors into individual data points of the same stage through a behavioral semantic mapping function, extended data units with interpretive constraints are formed and aggregated into stage-specific semantic constraint datasets with clear cognitive boundaries according to the exploration stage.

3. The geological and mineral exploration data processing method based on big data acquisition according to claim 2, characterized in that, In step S2, the specific indicators of cognitive consistency of data within the calculation phase include: For each data point in the staged semantic constraint dataset, extract its original set of observations and embedded behavioral semantic mapping, and obtain the relevant data set in its spatial or causal neighborhood; The consistency evaluation function is used to comprehensively calculate the consistency relationship between the original observations, behavioral semantic mapping and neighborhood data of the data, and output a numerical index that represents the degree of cognitive matching with the current stage.

4. The geological and mineral exploration data processing method based on big data acquisition according to claim 3, characterized in that, In step S2, dynamically generating the confidence interval specifically includes: The mean and standard deviation of the consistency index of stage cognition for all data within the current exploration stage are calculated. Based on the mean and standard deviation, and using the adjustment coefficient to control the expansion or convergence of the interval, the upper and lower bounds of the confidence interval representing the confidence level of the data at this stage are calculated.

5. The geological and mineral exploration data processing method based on big data acquisition according to claim 4, characterized in that, In step S2, the process of calculating the credible weight specifically includes: First, calculate the deviation of the stage cognition consistency index of each data point from the mean of the stage consistency index; Then, an exponential decay function is used to calculate the confidence weight of each data point based on the deviation, so that the data with a larger deviation receives a lower weight, thereby achieving a softening effect on abnormal data. Finally, the original observation data, behavioral semantic mapping, stage consistency index, credibility interval and credibility weight are integrated to form an enhanced data unit, which is then summarized into an enhanced stage credibility dataset.

6. The geological and mineral exploration data processing method based on big data acquisition according to claim 5, characterized in that, In step S3, dynamically reconstructing the network of relationships between data specifically includes: First, based on the credibility weights, spatial coordinates, and attribute vectors containing geophysical, geochemical, and engineering observations of the data units in the enhanced stage credibility dataset, the initial association strength between any two data units is calculated by combining the spatial distance function and the attribute similarity function. Subsequently, information reflecting the cognitive evolution of exploration stages, such as borehole verification results and geological interpretation model revisions, is introduced to generate relevant influencing factors. The initial association strength is dynamically modulated by the stage-based cognitive modulation coefficient to obtain the updated adaptive association strength.

7. The geological and mineral exploration data processing method based on big data acquisition according to claim 6, characterized in that, Step S3, specifically the adaptive sparsification of the interconnected network, includes: Based on the statistical distribution of all data in the current stage to the updated adaptive association strength, a stage adaptive threshold is dynamically determined to distinguish between strong and weak association edges. For edges identified as weakly correlated, a continuous edge weight decay strategy is adopted, which uses the edge weight decay coefficient to flexibly reduce the weight of weakly correlated edges.

8. The geological and mineral exploration data processing method based on big data acquisition according to claim 7, characterized in that, In step S4, identifying stable local structural units specifically includes: Based on the enhanced adaptive association dataset, a sparse association set between each data unit and other data units is extracted to construct a phased weighted association graph; Calculate the structural stability index for each data unit, which is related to the stability of the sum of the weights of its associated edges and the set of its neighborhood. Under the constraint of spatial continuity, multiple data units whose structural stability index exceeds the local structural stability threshold are aggregated to form a local structural unit.

9. A method for processing geological and mineral exploration data based on big data acquisition according to claim 8, characterized in that, In step S4, the adaptive determination of the expression granularity of each structural unit specifically includes: For each local structural unit, its granularity fit index is calculated. This index is obtained by combining the internal average confidence weight, internal attribute difference degree, and the mineral exploration task demand intensity index set by the engineering exploration plan of the structural unit through the granularity fit function. Based on the numerical range of the granularity adaptation index, an expression granularity level is adaptively selected for the structural unit from the original observation level, structural feature level, or target decision level.

10. A method for processing geological and mineral exploration data based on big data acquisition according to claim 9, characterized in that, In step S4, the process of generating standardized mineral exploration target representation units specifically includes: Based on the expression granularity level determined for each local structural unit, the corresponding feature mapping function is used to map the features of each data unit within that structural unit; Based on the comprehensive contribution weight of each data unit, the mapped features are weighted and fused to generate the granular adaptive expression vector of the structural unit. Finally, the granular adaptive expression vector, the spatial coverage information of the structural unit, and the overall credibility are integrated to construct a standardized mineral exploration target expression unit, and summarized to form a stage mineral exploration target expression set.

Images