Multi-source data fusion based whole-process digital management and control system for geological exploration enterprises

Abstract

The application discloses a multi-source data fusion-based whole-process digitalized management and control system for geological exploration enterprises, and particularly relates to the technical field of information processing in the geological exploration industry, and comprises the following steps: a multi-source data fusion module loads a template and geological rules according to a project type, generates a fusion data body by performing space-time alignment and semantic verification on multi-source exploration data, a field closed-loop management and control module constructs a risk situation map according to the fusion data body and a real-time state, generates a task scheme and issues the task scheme, and realizes feedback through a closed loop, a knowledge graph evaluation module constructs a graph neural network according to a geological knowledge graph, inputs a fusion data raster, and generates an evaluation result, a three-dimensional collaborative modeling module constructs and updates a geological interface model according to the fusion data and the feedback, and carries the evaluation result, and a strategic cockpit module generates a scheduling scheme in combination with a resource state according to the evaluation result and the updated model, and calculates a value change by calling an economic evaluation model. The system realizes closed-loop linkage and whole-process digitalized management and control of geological exploration project data, construction and decision-making.

Description

Technical Field

[0001] This invention relates to the field of information processing technology in the geological exploration industry, and more specifically, to a digital and intelligent management and control system for the entire process of geological exploration enterprises based on multi-source data fusion. Background Technology

[0002] Geological exploration companies are involved in the acquisition and processing of heterogeneous data from multiple sources, including geological surveying, geophysical exploration, geochemical exploration, remote sensing, drilling, in-situ testing, and laboratory experiments. The data formats encompass various types, such as structured tables, semi-structured documents, unstructured images and point clouds, and time-series sensor data. At the operational level, geological exploration projects require the management of the entire process from project initiation, fieldwork, data processing to results archiving, covering multiple dimensions such as technical management, schedule control, quality control, cost accounting, and resource allocation. With the development of information technology, some companies have introduced project management software, geological data platforms, 3D modeling tools, and laboratory information management systems to attempt to achieve information-based management of their geological exploration operations.

[0003] However, in practical use, it still has some shortcomings. For example, multi-source data fusion is mostly limited to format conversion and coordinate unification, lacking geoscientific semantic logic verification, making it difficult to automatically identify recording errors, stratigraphic contradictions, and conflicts between geophysical and drilling results, resulting in insufficient data credibility. Field construction management relies on manual reporting and regular inspections, lacking a closed-loop mechanism from on-site status perception to automatic task reordering. Exploration target evaluation relies on personal experience, and geoscientific expert knowledge is not structured and integrated into quantitative evaluation models, resulting in insufficient interpretability and repeatability. Three-dimensional geological models are mostly constructed once and are difficult to update dynamically as construction progresses, and offline collaborative plotting lacks conflict resolution mechanisms. Exploration results and business decisions are disconnected, and geological knowledge updates cannot automatically drive resource scheduling and value assessment. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, this invention provides a digital and intelligent management and control system for the entire process of geological exploration enterprises based on multi-source data fusion, which solves the problems mentioned in the background art through the following solutions.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a digital and intelligent management and control system for the entire process of geological exploration enterprises based on multi-source data fusion, comprising: The multi-source data fusion module is used to load the corresponding data template and geological rule set according to the business type of the created project, perform spatiotemporal alignment and semantic verification on the exploration data collected from multiple sources, and generate the standardized fusion data body of the project. The on-site closed-loop management module is used to acquire the real-time construction status synchronized from the exploration site, construct the geological risk situation map of the current stage of the project based on the standardized fusion data and the real-time construction status, generate an exploration task scheduling scheme adapted to the current working conditions, and send the exploration task scheduling scheme to the execution end. The knowledge graph evaluation module is used to call the corresponding geoscience knowledge graph subset according to the project type, and construct a graph neural network based on the geoscience knowledge graph subset; and to rasterize the standardized fused data volume into spatial features, input them into the graph neural network, perform quantitative evaluation of the exploration target, and generate exploration target evaluation results. The three-dimensional collaborative modeling module is used to construct and dynamically update a general geological interface model based on the standardized fused data volume and the execution feedback data of the exploration task scheduling scheme; the geological interface model carries the evaluation results of the exploration target and forms the updated geological interface model; The strategic cockpit module is used to dynamically calculate the changes in project value by calling an economic evaluation model that matches the project type based on the evaluation results of the exploration target and the updated geological interface model, and to perform global optimization in combination with resource status to generate a scheduling plan for personnel and equipment.

[0006] The technical effects and advantages of this invention are as follows: 1. By driving the loading of data templates and geological rule sets through business types, and performing spatiotemporal alignment and geoscientific semantic verification on multi-source exploration data, it automatically identifies data quality issues such as inverted logging logic, contradictions between adjacent projects, and conflicts between geophysical and drilling results, and generates standardized fused data bodies with embedded quality tags, providing a reliable data foundation for full-process management; 2. By jointly analyzing real-time construction status with standardized integrated data, calling the geological risk causal diagram in the form of a dynamic Bayesian network for risk reasoning, automatically generating and issuing exploration task scheduling plans, and then continuously and dynamically adjusting through execution feedback, a closed-loop management system is formed, which elevates field construction management from post-event traceability to pre-event prediction and real-time response. 3. By initializing and constraining the graph neural network with the topological structure and quantitative constraint rules of a subset of the geoscience knowledge graph, expert knowledge is embedded into the exploration target evaluation model in a structured manner, making the evaluation process geoscientific interpretable and the evaluation results traceable. This avoids the "black box" output problem of traditional deep learning models and improves the accuracy and credibility of quantitative evaluation. 4. By implementing a feedback data-driven local radial basis function incremental update mechanism, the three-dimensional geological interface model can continuously evolve as construction progresses without the need for full-area reconstruction; at the same time, by using an operation transformation algorithm to perform conflict detection and intelligent resolution on offline collaborative plotting, the consistency problem of collaborative interpretation by multiple geological personnel in a weak network environment is solved. 5. By automatically integrating the knowledge graph evaluation results with the dynamically updated geological interface model into the economic evaluation model that matches the project type, the project value changes are dynamically calculated. Based on this, the multi-constraint resource scheduling optimization is driven by demand, realizing an automated closed loop from geological knowledge updates to business decision adjustments, and solving the problem of the separation between exploration results and business decisions. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0008] Figure 2 This is a schematic diagram of the multi-source data fusion structure of the present invention.

[0009] Figure 3 This is a schematic diagram of the on-site closed-loop control structure of the present invention.

[0010] Figure 4 This is a schematic diagram of the knowledge graph evaluation and three-dimensional collaborative modeling structure of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] refer to Figures 1-4 The digital and intelligent management and control system for the entire process of geological exploration enterprises, as shown, includes: The multi-source data fusion module is used to load corresponding data templates and geological rule sets according to the business type of the created project, perform spatiotemporal alignment and semantic verification on exploration data collected from multiple sources, and generate a standardized fused data body for the project. In the full-process digital and intelligent management and control system of geological exploration enterprises, when a new exploration project is created, the user first specifies the business type of the project, and the system then retrieves the data template and geological rule set corresponding to the business type from the pre-set configuration library. After loading, the module receives exploration data from multiple sources through the multi-protocol acquisition adapter, synchronously records the source identifier, acquisition timestamp and device identifier of each data, and converts the heterogeneous source field names into system standard field names according to the preset field mapping relationship in the data template. After data acquisition and field mapping are completed, the module performs spatiotemporal alignment and semantic verification on the aggregated exploration data in sequence. Spatiotemporal alignment unifies data with different spatial reference systems and time bases into the same coordinate frame and time base. Semantic verification uses the loaded geological rule set to automatically check the rationality and consistency of the data, adds verification result labels to each data and records data quality events.

[0013] After the above processing, the module integrates the multi-source exploration data that has undergone spatiotemporal alignment and semantic verification according to the structure specified in the data template, generating a standardized fused data body organized by project. The standardized fused data body is divided into multiple data layers according to data type. The data layers are linked by spatial coordinates, timestamps, and project identifiers, and data quality tags and data quality event records generated during the semantic verification stage are embedded. The standardized fused data body is stored in the system data platform and is provided to the outside world in the form of service interfaces. Subsequent updates and supplements to the data by each module are also written back to the standardized fused data body through update interfaces.

[0014] Furthermore, the specific execution process of the spatiotemporal alignment includes: Spatially, the module identifies the spatial location information in each data entry, expressed in forms such as latitude and longitude coordinates, projected coordinates, engineering relative coordinates, borehole opening coordinates and depth descriptions, and survey line station numbers. A built-in spatial coordinate transformation engine uniformly converts these spatial locations to a representation under the project's preset target spatial reference system. For location information with only textual descriptions, the module uses geocoding technology, combined with a digital map and place name database of the project area, to convert it into vector graphics with spatial extent. Temporally, the module uniformly converts the timestamps of each data source to Coordinated Universal Time (UTC) format, records the original time zone offset, and for project data spanning multiple surveys or construction phases, it stages the data according to project phases, establishing a time-series index for the data.

[0015] Furthermore, the specific execution process of the semantic verification includes: for lithology, age, and stratigraphic records of the same borehole or profile, the module checks for logical inversions where new strata appear beneath older strata according to stratigraphic superposition rules; if such inversions occur, they are marked as abnormal and a warning is given that they may be caused by cataloging errors or tectonic inversions; for multiple spatially adjacent exploration points, the module checks for contradictions between stratigraphic divisions and lithological descriptions according to regional geological background rules; if such contradictions exist, they are marked as abnormal; for geophysical interpretation profiles and drilling results within the same region, the module compares the stratigraphic boundary depths revealed by the two types of data; if the difference exceeds a preset tolerance range, it is marked as a conflict point and a side-by-side comparison view is provided; for experimental test data, the module scans for numerical points exceeding the reasonable range according to the reasonable value range rules for test indicators and marks them accordingly. Verification result labels include Pass, Abnormal (requires review), and Conflict (pending adjudication).

[0016] Furthermore, the construction process of the standardized fusion data body includes: the module classifies the data into four layers according to type: basic geographic layer, exploration engineering layer, testing and analysis layer, interpretation results layer, and time-series monitoring layer. The basic geographic layer stores spatial framework data such as topography, imagery, and project boundaries; the exploration engineering layer stores the location, trajectory, and attribute information of exploration projects such as boreholes, pits, trenches, and survey lines; the testing and analysis layer stores sample test results, in-situ test data, and physical property parameters; the interpretation results layer stores interpretation results such as geological profiles, geological patches, and anomaly zone delineation; and the time-series monitoring layer stores monitoring time-series data such as water level, deformation, and stress. The layers are linked through unified spatial coordinates, timestamps, and engineering identifiers.

[0017] The on-site closed-loop management module is used to acquire real-time construction status synchronized from the exploration site, construct a geological risk situation map of the current stage of the project based on the standardized fused data and the real-time construction status, generate an exploration task scheduling scheme adapted to the current working conditions, and distribute the exploration task scheduling scheme to the execution terminal.

[0018] After the multi-source data fusion module generates a standardized fused data body, the on-site closed-loop management module starts operating. The module first acquires the real-time construction status synchronized from the exploration site. The real-time construction status is synchronously transmitted back through sensors, terminal equipment deployed at the exploration site, and manual reporting channels, reflecting the construction progress and physical conditions of each exploration project at the current moment. After these data are uniformly accessed through a multi-protocol acquisition adapter, they are associated with the exploration project layer data in the standardized fused data body through project identifiers and timestamps. After acquiring the real-time construction status, the module performs joint analysis with the standardized fusion data to construct a geological risk situation map for the current stage of the project. Based on this geological risk situation map, the module further generates an exploration task scheduling scheme adapted to the current working conditions. The system extracts construction points and areas with risk levels exceeding the preset threshold from the situation map, matches corresponding countermeasures according to the risk type, and generates regular continuous tasks for construction points with normal risk levels according to the schedule. Then, it sorts and resolves conflicts globally based on the pre-dependencies between tasks, resource allocation constraints, and construction space conflict rules to form a complete exploration task arrangement scheme. After the plan is generated, the module distributes and pushes it to each execution terminal according to the role and permissions of the task executor. During the execution of the plan, the module continuously collects execution feedback data and compares it with the issued plan item by item. When a task is not completed as planned, the construction status changes significantly, or the geological risk situation map is updated and shows new high-risk points, the dynamic adjustment of the exploration task arrangement plan is triggered, the adjusted plan is generated and reissued, forming a continuous closed-loop management.

[0019] Furthermore, the specific execution process for constructing the geological risk situation map includes: first, comparing the time-series data of drilling parameters in the real-time construction status with the existing geological profiles and stratigraphic information in the standardized fusion data body to identify abnormal working conditions such as sudden torque changes, abnormal fluctuations in drilling speed, and sudden increases in mud loss during drilling, and spatially associating these abnormal working conditions with the "abnormal and needing verification" data points marked in the semantic verification stage of the standardized fusion data body; then, calling a pre-set geological risk causal graph model, which is expressed in the form of a dynamic Bayesian network, whose nodes include the degree of stratigraphic fracturing, the degree of fracture development, the state of groundwater occurrence, the probability of borehole deviation, the probability of stuck drill bit, and the probability of borehole wall collapse, etc., and the directed edges between nodes represent the causal relationship and its probability strength; inputting the real-time construction status characteristics of each construction point into the model, updating the posterior probability of each risk node through probabilistic inference, and obtaining the main risk types currently faced by each construction point and their probability of occurrence; finally, overlaying and rendering the risk information onto the project geographic base map in the form of heat maps, risk level symbols, risk labels, etc., to generate a geological risk situation map.

[0020] Furthermore, when generating the exploration task scheduling plan, the specific methods for matching response measures according to the risk type include: generating drilling parameter adjustment tasks such as reducing drilling pressure and increasing mud weight for boreholes with high risk of stuck drill bit; generating quality control tasks such as increasing the frequency of deviation measurement and pre-setting deviation correction schemes for boreholes with high risk of deviation; generating construction measures such as supplementing casing and adjusting drilling technology for boreholes that expose fracture zones; and generating resource allocation tasks such as dispatching additional units and adjusting shifts for boreholes with serious delays.

[0021] Furthermore, the specific methods for distributing the exploration task scheduling plan to the execution end include: the module pushes drilling parameter adjustment instructions and shift work plans to drilling rig operators according to roles and permissions, pushes sampling requirements and logging tasks to geological logging personnel, pushes sample delivery notices and testing tasks to testing and inspection personnel, and pushes a global task scheduling view and key risk warnings to project management personnel; the execution end includes the industrial control display screen deployed on the drilling rig, the handheld terminals of on-site personnel, and the monitoring terminal of the project management department.

[0022] The knowledge graph evaluation module is used to call the corresponding geoscience knowledge graph subset according to the project type, and construct a graph neural network based on the geoscience knowledge graph subset; it rasterizes the standardized fused data volume into spatial features, inputs them into the graph neural network, performs quantitative evaluation of the exploration target, and generates exploration target evaluation results.

[0023] After the multi-source data fusion module generates a standardized fused data volume, the knowledge graph evaluation module begins operation. The module first retrieves a subset of geoscientific knowledge graphs corresponding to the business type determined during project creation from the system's pre-built geoscientific knowledge graph repository. Based on this subset, the module constructs a graph neural network. Using concept nodes and semantic relation edges from a subset of the geoscience knowledge graph as the initial nodes and edges of the graph neural network, the topology of the knowledge graph is directly mapped to the network skeleton. The quantitative constraint rules and association probabilities stored in the graph are transformed into the initial weights of the corresponding edges and the parameter constraints in the message passing function. Then, the network is fine-tuned and trained using known geological data and verified evaluation conclusions of the project area. While constructing the graph neural network, the module rasterizes the standardized fused data volume into spatial features: within the evaluation area, a regular grid is divided according to a preset spatial resolution; exploration data falling into each grid is extracted from each data layer of the standardized fused data volume, and a multi-dimensional spatial feature vector of each grid cell is calculated; the module inputs the rasterized spatial features into the graph neural network, with each grid cell mapped to an input node, and the spatial adjacency relationship between grids mapped to edges; concept nodes in the geoscience knowledge graph subset are also connected to the corresponding grid nodes; through multi-layer message passing and feature aggregation, a quantitative evaluation value is output for each grid cell; finally, the module summarizes the quantitative evaluation values ​​of each grid cell to generate an exploration target evaluation result containing spatial location information, evaluation level, and confidence level.

[0024] Furthermore, the specific methods for calling up subsets of the geoscience knowledge graph include: the system pre-dividing the overall geoscience knowledge graph database into multiple subsets according to business type. The mineral exploration subset includes concepts and relationships such as tectonic mineralization, magmatic mineralization, alteration zoning, mineralization types, and geophysical and geochemical anomaly interpretation; the engineering exploration subset includes concepts and relationships such as foundation soil types, engineering geological layer division, karst development patterns, fault activity, and seismic effects; and the disaster geology subset includes concepts and relationships such as landslide formation conditions, debris flow source characteristics, collapse structural surfaces, and ground subsidence mechanisms. Each subset also contains quantitative constraint rules, such as the correlation probability between specific alteration mineral combinations and mineralization intensity, the probability of karst formation at a specific burial depth for a specific lithological combination, and the probability range of landslide occurrence under specific terrain slope and rainfall conditions.

[0025] Furthermore, the specific execution process of rasterizing the standardized fused data volume into spatial features includes: the module first determines the boundary range of the evaluation area, which is a certain buffer distance extending outward from the boundary of the project exploration area, and then divides the area into regular rectangular or hexagonal grids according to a preset spatial resolution; for each grid cell, geochemical features such as geochemical element anomaly intensity, element combination index, and element ratio are extracted from the test analysis layer; engineering exposure features such as borehole mineralization depth, mineralization thickness, and mineralization occurrence are extracted from the exploration engineering layer; features such as geochemical anomaly amplitude, remote sensing alteration information intensity, structural line density, and direction are extracted from the interpretation results layer; and topographic features such as terrain slope, elevation, and distance to water systems are extracted from the basic geographic layer; for sparse grids, spatial interpolation methods are used to estimate and fill the data using known data from neighboring grids.

[0026] Furthermore, the specific meaning of the quantitative evaluation value output by the graph neural network varies depending on the business type: mineral exploration projects output mineralization favorability, representing the probability of mineralization or expected mineralization intensity in that grid cell; engineering exploration projects output the suitability level of engineering geological conditions or the probability of development of unfavorable geological bodies; and disaster geology projects output the geological hazard risk index or instability probability. During training and inference, the quantitative constraint rules stored in the geoscience knowledge graph subset constrain the network output through the regularization term in the loss function.

[0027] The 3D collaborative modeling module is used to construct and dynamically update a general geological interface model based on the standardized fused data volume and the execution feedback data of the exploration task scheduling scheme. The geological interface model carries the evaluation results of the exploration targets, forming an updated geological interface model.

[0028] Based on the output data from the multi-source data fusion module and the on-site closed-loop management module, the 3D collaborative modeling module begins operation; the module first acquires two types of basic data required for building and updating the model: The first category includes the location, trajectory, and layered description of drilling, pitting, trenching, survey lines, and other engineering works in the standardized fusion data body, as well as the geological profiles and geological patches of the interpreted result layers; The second category is the execution feedback data collected by the on-site closed-loop control module during the execution of the exploration task scheduling plan, including newly completed borehole logging columnar sections, newly acquired in-situ test curves, newly revealed stratigraphic boundary depths, and newly discovered lithological changes or tectonic signs. Based on the above data, a general geological interface model is constructed using an implicit modeling method: The spatial coordinates of the stratigraphic interface points exposed by each exploration project are extracted as known control points. A scalar field is constructed in the three-dimensional space of the entire exploration area using the radial basis function interpolation algorithm. Isosurfaces with the same identifier value are used as the corresponding geological interfaces, and geological profiles and geological patches are introduced as trend constraints. As exploration and construction progress, the module receives new execution feedback data, extracts newly added stratigraphic interface control points, and uses a local radial basis function update algorithm to recalculate the scalar field and adjust the isosurface only within the influence radius of the newly added control points, achieving dynamic model updates. During model construction and updating, the module also supports offline collaborative plotting of the same geological interface model by multiple clients. The module maps the generated exploration target evaluation results to the corresponding locations on the geological interface model through spatial coordinate registration, ultimately forming the updated geological interface model.

[0029] Furthermore, the specific process of constructing a general geological interface model using the implicit modeling method includes: each control point carries a label of its respective stratigraphic interface type, and the value of the scalar field at each control point is equal to the preset identifier value of the stratigraphic interface to which that point belongs; for complex geological phenomena such as stratigraphic pinch-outs, lenses, and fault cuts, geological profiles and geological patches of the interpreted strata are introduced as trend constraints during the interpolation process, so that the generated isosurfaces can still maintain a reasonable geological morphology in areas far from the control points.

[0030] Furthermore, the specific execution process of the dynamic update includes: after receiving the execution feedback data, the module extracts the newly revealed stratigraphic interface points and their spatial coordinates and stratigraphic labels, and adds these new control points to the existing control point set; then, a local radial basis function update algorithm is used to recalculate the scalar field and adjust the isosurface only within the influence radius of the new control points; the influence radius is determined by the spatial distribution density of the control points and the spatial variability parameters of the stratigraphy; the updated model strictly passes through the newly revealed stratigraphic interface at the new control points, and maintains the original model shape in areas far away from the new points.

[0031] Furthermore, the specific execution process of the offline collaborative mapping includes: before entering the work area, geological personnel from each region download the model geometric tiles and attribute data within the current work area from the server using handheld terminals; in the offline state, they perform mapping operations on the model, and the drawn mapping lines are automatically snapped onto the surface of the three-dimensional geological interface model and bound to the stratigraphic semantic attributes and operator identity information; when the terminal restores the network connection, it uploads the mapping operation records during the offline period to the server; after receiving the mapping operations uploaded by multiple terminals, the server uses an operation conversion algorithm to perform weighted fusion on multiple mappings that overlap or are close in space, based on the historical operation consistency weight of each operator, the geometric proximity of the mappings, and the stratigraphic semantic compatibility, to generate continuous geological boundaries and mark them as pending review status, while retaining each original mapping version and resolution process record.

[0032] Furthermore, the specific method for carrying the evaluation results of the exploration target is as follows: the quantitative evaluation value expressed in the form of grid cells in the evaluation results of the exploration target is mapped to the corresponding position of the geological interface model through spatial coordinate registration, so that the evaluation value of each grid cell is stored and displayed as an additional attribute of the corresponding position of the geological interface model below it.

[0033] The strategic cockpit module is used to dynamically calculate the changes in project value based on the exploration target evaluation results and the updated geological interface model, by calling an economic evaluation model that matches the project type, and to perform global optimization in conjunction with resource status, generating a scheduling plan for personnel and equipment.

[0034] The strategic cockpit module runs after the knowledge graph evaluation module and the 3D collaborative modeling module output their results. The module first receives the exploration target evaluation results and the updated geological interface model. Based on these two inputs, it calls an economic evaluation model that matches the project's business type to dynamically calculate changes in project value. After completing the project value change calculation, the module performs global optimization based on resource status, generating a personnel and equipment scheduling plan. The supplementary or adjusted exploration tasks identified in the value change calculation are treated as new requirements and input into the resource scheduling optimizer. With the goal of minimizing the total scheduling cost and maximizing the key task satisfaction rate, the optimizer searches and generates scheduling schemes for personnel and equipment while satisfying constraints such as equipment capacity, personnel skill matching, construction space conflict, and task time window. The scheduling plan clearly indicates the dispatching, relocation, task content, time schedule and relocation route of each unit and personnel. After the scheduling plan is generated, it is pushed to the enterprise management in a visual form. After confirmation, it is distributed to the project execution end and the unit's handheld terminal, which actually drives the redeployment of resources.

[0035] Furthermore, the specific execution process of dynamically calculating the changes in project value includes: for mineral exploration projects, invoking the resource quantity-net present value model. This model is triggered each time an updated geological interface model and exploration target evaluation results are received. Based on the spatial distribution changes of the ore body interface in the updated geological interface model and the updated mineralization favorability level in the exploration target evaluation results, the ore quantity and metal quantity of each ore body block are re-estimated. Combined with the mineral product market price forecast curve, mining, beneficiation, and metallurgical technical and economic parameters, and discount rate, the current net present value and internal rate of return of the project are calculated. The project's value is dynamically changed by comparing the difference with the historical value obtained from the previous calculation. For engineering survey projects, the quantity-profit margin model is invoked. This model re-estimates the number of boreholes to be added or reduced, the number of in-situ test points, and the number of indoor test groups based on the updated geological interface model, the changes in soil and rock stratification revealed in the updated geological interface model, and the updated probability of the development of adverse geological bodies in the evaluation results of the exploration targets. Combined with the contract unit price and cost parameters, the project's current expected profit margin is calculated and compared with the historical expected profit margin obtained from the previous calculation, outputting the dynamic change of the expected profit margin.

[0036] Furthermore, the specific constraints of the resource scheduling optimizer include: equipment capacity constraints, which refer to the restrictions on the types of executable tasks imposed by each drilling rig model and drilling capacity, and the working frequency and technical parameters of each geophysical equipment; personnel skill matching constraints, which refer to the matching requirements between the skill tags and qualification levels of each technician and the skills required for the task; construction space conflict constraints, which refer to the safety distance and construction sequence restrictions for multiple devices operating simultaneously within the same construction area; and task time window constraints, which refer to the earliest start time, latest completion time, and schedule requirements for each task.

[0037] Furthermore, the visualization and push methods for the scheduling scheme include: displaying a global resource distribution map and a scheduling Gantt chart on the large screen interface of the strategic cockpit, and sending scheduling suggestions to the terminals of relevant management personnel in the form of message notifications. The module also records and reviews the execution status and effects of each scheduling, continuously compares and analyzes changes in project value with actual resource input, and provides feedback for the parameter calibration of subsequent economic evaluation models.

[0038] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A digital and intelligent management and control system for the entire process of geological exploration enterprises based on multi-source data fusion, characterized in that: include: The multi-source data fusion module is used to load the corresponding data template and geological rule set according to the business type of the created project, perform spatiotemporal alignment and semantic verification on the exploration data collected from multiple sources, and generate the standardized fusion data body of the project. The on-site closed-loop management module is used to acquire the real-time construction status synchronized from the exploration site, construct the geological risk situation map of the current stage of the project based on the standardized fusion data and the real-time construction status, generate an exploration task scheduling scheme adapted to the current working conditions, and send the exploration task scheduling scheme to the execution end. The knowledge graph evaluation module is used to call the corresponding geoscience knowledge graph subset according to the project type, and construct a graph neural network based on the geoscience knowledge graph subset; and to rasterize the standardized fused data volume into spatial features, input them into the graph neural network, perform quantitative evaluation of the exploration target, and generate exploration target evaluation results. The three-dimensional collaborative modeling module is used to construct and dynamically update a general geological interface model based on the standardized fused data volume and the execution feedback data of the exploration task scheduling scheme; the geological interface model carries the evaluation results of the exploration target and forms the updated geological interface model; The strategic cockpit module is used to dynamically calculate the changes in project value by calling an economic evaluation model that matches the project type based on the evaluation results of the exploration target and the updated geological interface model, and to perform global optimization in combination with resource status to generate a scheduling plan for personnel and equipment.

2. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The construction of the standardized fusion data body includes: The multi-source exploration data, after spatiotemporal alignment and semantic verification, are categorized according to the data template and data type into four layers: basic geographic layer, exploration engineering layer, testing and analysis layer, interpretation results layer, and time-series monitoring layer. The basic geographic layer stores topographic data, imagery, and project boundary spatial framework data; the exploration engineering layer stores the location, trajectory, and attribute information of boreholes, pits, trenches, and survey lines; the testing and analysis layer stores sample test results, in-situ test data, and physical property parameters; the interpretation results layer stores geological profiles, geological patches, and anomaly zone delineation interpretation results; and the time-series monitoring layer stores water level, deformation, and stress monitoring time-series data. Each data layer is linked through spatial coordinates, timestamps, and engineering identifiers. Data quality tags and data quality event records generated during the semantic verification phase are embedded into the standardized fused data body.

3. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The construction of the geological risk situation map includes: First, the time-series data of drilling parameters in real-time construction status are compared with the existing geological profiles and stratigraphic information in the standardized fusion data body to identify abnormal working conditions such as sudden torque changes, abnormal fluctuations in drilling speed, and a sudden increase in mud loss during drilling. These abnormal working conditions are then spatially correlated with the abnormal data points marked in the semantic verification stage of the standardized fusion data body that require review. Next, a geological risk causal graph model expressed in the form of a dynamic Bayesian network is invoked, and the real-time construction status characteristics of each construction point are input into the model. The posterior probability of each risk node is updated through probabilistic inference to obtain the main risk types currently faced by each construction point and their probability of occurrence. Finally, the risk information is overlaid and rendered on the project's geographical base map in the form of a heat map, risk level symbols, and risk labels to generate the geological risk situation map.

4. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The generation of the exploration task scheduling scheme includes: Construction points and areas with risk levels exceeding preset thresholds are extracted from the geological risk situation map. Countermeasures are matched according to risk type. For boreholes with high risk of stuck drill pipe, drilling parameter adjustment tasks are generated, such as reducing drilling pressure and increasing mud weight. For boreholes with high deviation risk, quality control tasks are generated, such as increasing the frequency of deviation measurements and pre-set deviation correction schemes. For boreholes revealing fractured zones, construction measures tasks are generated, such as adding casing and adjusting drilling technology. For boreholes with severely delayed progress, resource allocation tasks are generated, such as dispatching additional rigs and adjusting shifts. Simultaneously, for construction points with risk levels within the normal range, routine drilling, sampling, logging, and testing tasks are generated according to the overall project schedule and standard construction procedures. Then, based on the pre-determined dependencies between tasks, resource allocation constraints, and construction space conflict rules, a global sorting and conflict resolution process is performed to form the exploration task arrangement scheme.

5. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The spatial features include: Geochemical characteristics including geochemical element anomaly intensity, element combination index, and element ratio extracted from the test analysis layer; engineering exposure characteristics including borehole depth of mineralization, mineralization thickness, and mineralization occurrence extracted from the exploration engineering layer; geophysical and remote sensing characteristics including geophysical anomaly amplitude, remote sensing alteration information intensity, and structural line density and direction extracted from the interpretation results layer; topographic and geomorphological characteristics including topographic slope, elevation, and distance to drainage systems extracted from the basic geographic layer; and for sparse grids, estimation and filling are performed using known data from neighboring grids through spatial interpolation methods.

6. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The generation of the exploration target evaluation results includes: The spatial features obtained by rasterization are input into the graph neural network. Each grid cell is mapped to an input node, and the spatial adjacency between grid cells is mapped to edges. Concept nodes related to the evaluation task in the geoscience knowledge graph subset are connected to the corresponding grid nodes. Through multi-layer message passing and feature aggregation, a quantitative evaluation value is output for each grid cell. Among them, mineral exploration projects output the mineralization favorability, engineering exploration projects output the engineering geological condition suitability level or the probability of development of adverse geological bodies, and disaster geology projects output the geological disaster risk index or the probability of instability. The quantitative evaluation values ​​of each grid cell are summarized to generate an exploration target evaluation result containing spatial location information, evaluation level, and confidence level. The evaluation level is divided into favorable area, relatively favorable area, and unfavorable area according to a preset threshold range.

7. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The construction and dynamic updating of the general geological interface model include: The spatial coordinates of stratigraphic interface points revealed by each exploration project in the standardized fusion data volume are extracted as known control points. Each control point carries a label of its corresponding stratigraphic interface type. A scalar field is constructed in the three-dimensional space of the exploration area using a radial basis function interpolation algorithm. Isosurfaces with the same label value are used as the corresponding geological interfaces. During the interpolation process, geological profiles and geological patches of the interpreted strata in the standardized fusion data volume are introduced as trend constraints. After receiving the execution feedback data, newly added stratigraphic interface points, their spatial coordinates, and stratigraphic labels are extracted. The newly added control points are added to the existing control point set. A local radial basis function update algorithm is used to recalculate the scalar field and adjust the isosurfaces only within the influence radius of the newly added control points. The influence radius is determined by the spatial distribution density of the control points and the spatial variability parameters of the strata.

8. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The carrier of the exploration target evaluation results includes: The quantitative evaluation values ​​expressed in grid cells in the exploration target evaluation results are mapped to the corresponding positions on the geological interface model through spatial coordinate registration. The evaluation value of each grid cell is stored and displayed as an additional attribute of the corresponding position on the geological interface model below it. When the exploration target evaluation result is the mineralization favorability of mineral exploration, it carries the mineralization favorability level and confidence level of each grid cell. When the evaluation result is the engineering geological condition suitability level of engineering exploration, it carries the suitability level and probability of development of adverse geological bodies of each grid cell.

9. The intelligent digital management and control system for the entire process of geological exploration enterprises based on multi-source data fusion as described in claim 1, characterized in that, The calculation of the change in the project value includes: The economic evaluation models include: resource quantity-net present value model and project quantity-profit rate model; For mineral exploration projects, the resource quantity-net present value model is invoked and is triggered each time an updated geological interface model and exploration target evaluation results are received. Based on the spatial distribution changes of the ore body interface in the updated geological interface model and the update of the mineralization favorability level in the exploration target evaluation results, the ore quantity and metal quantity of each ore body block are re-estimated. Combining the mineral product market price forecast curve, mining, beneficiation and metallurgical technical and economic parameters and discount rate, the current net present value and internal rate of return of the project are calculated. The difference between these values ​​and the historical values ​​obtained from the previous calculation is compared, and the dynamic change in the project value is output. For engineering survey projects, the quantity-profit margin model is invoked. Based on the updated geological interface model revealing changes in soil and rock stratification and the updated probability of adverse geological body development in the exploration target evaluation results, the number of additional or reduced boreholes, in-situ test points, and indoor test groups are re-estimated. Combined with the contract unit price and cost parameters, the current expected profit margin of the project is calculated. This is then compared with the historical expected profit margin calculated previously, and the dynamic change in the expected profit margin is output.

Images