Geological exploration abnormal signal identification and communication triggering system based on knowledge graph

Abstract

The invention relates to the technical field of geological exploration, in particular to a geological exploration abnormal signal identification and communication triggering system based on a knowledge graph, which periodically updates an entity association sequence of each exploration device, can timely screen out a core strong association entity and eliminate invalid association, and improves the accuracy of the geological exploration abnormal signal identification and communication triggering. It is ensured that anomaly analysis always focuses on entities strongly related to equipment services, and identification deviation caused by irrelevant entity interference is avoided; meanwhile, all associated entities do not need to be traversed in each anomaly analysis, redundancy calculation is reduced, and the dual requirements of geological exploration anomaly recognition for precision and efficiency are effectively balanced; according to the method, the number of the abnormal devices and the association strength between the devices are quantified into the communication index, and triggering objects of different permissions are matched through the index range, so that accurate correspondence of'abnormal severity-disposal subject permission 'is realized, waste of high-permission resources occupied by slight exceptions is avoided, and disposal lag caused by low-permission response of major exceptions is also prevented.

Description

Technical Field

[0001] This invention relates to the field of geological exploration technology, and more specifically, to a knowledge graph-based system for identifying and triggering geological exploration anomaly signals. Background Technology

[0002] In the field of geological exploration technology, traditional geological data management and knowledge integration largely rely on scattered structured tables, semi-structured reports, and unstructured documents, making it difficult to achieve deep integration of multi-source data such as stratigraphic exploration data, equipment operation logs, and expert experience texts. More importantly, the relationships between entities such as strata, equipment, and signals in traditional methods are mostly statically preset and cannot be dynamically adjusted as new exploration data accumulates and geological understanding deepens. This results in entity relationships failing to align with the evolution of actual exploration operations, affecting the fundamental reliability of subsequent anomaly signal analysis and hindering the efficient extraction of hidden geological correlation information.

[0003] In the identification of abnormal signals from exploration equipment, traditional technologies lack a systematic screening and updating mechanism for entities associated with the equipment. In most cases, anomaly analysis requires traversing all entity data related to the equipment. This includes a large amount of redundant information with low business relevance, which is prone to identification bias due to interference from irrelevant entities. Furthermore, since related entities are not sorted according to business importance, each anomaly determination requires reprocessing the entire dataset, increasing computational costs and reducing anomaly identification response efficiency. It is difficult to balance the dual requirements of geological exploration for identification accuracy and real-time performance.

[0004] Traditional methods for triggering communication after anomalies often rely on manual judgment of the anomaly severity to determine the notification recipients, lacking quantitative standards and exhibiting strong subjectivity. Based on this, this application proposes a knowledge graph-based system for identifying and triggering communication of geological exploration anomaly signals. Summary of the Invention

[0005] In view of the shortcomings of existing technologies, the purpose of this invention is to provide a geological exploration anomaly signal identification and communication triggering system based on knowledge graphs.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A knowledge graph-based geological exploration anomaly signal identification and communication triggering system includes:

[0008] The geological exploration knowledge graph construction module is used to construct a knowledge graph in the field of geological exploration.

[0009] The entity association sequence update module updates the entity association sequence of each exploration device regularly based on the knowledge graph in the geological exploration field.

[0010] The exploration anomaly signal identification module is used to determine whether there are exploration signal anomalies in each exploration device. If there are no exploration signal anomalies in all exploration devices, communication does not need to be triggered. If at least one exploration device has an exploration signal anomaly, the geological exploration communication index is calculated.

[0011] The abnormal communication triggering module determines the communication triggering object based on the geological exploration communication index and sends abnormal exploration signals to the corresponding communication triggering object.

[0012] Furthermore, the technical framework of knowledge graphs in the field of geological exploration adopts a six-level progressive architecture: "data layer → extraction layer → fusion layer → storage layer → reasoning layer → update layer." The extraction layer's implementation steps include: extracting knowledge triples, and then... Dynamically quantify entity correlation; among which, , , All are weighting coefficients. ; The number of times an entity co-occurs; For entities Total number of occurrences; For entity semantic similarity; Dependence on the physical domain.

[0013] Furthermore, the update process of an exploration equipment's entity association sequence is as follows: Select an exploration equipment, determine the entity position of the exploration equipment in the geological exploration domain knowledge graph, mark all entities in the geological exploration domain knowledge graph that are related to the entity but do not belong to attribute information as associated entities, obtain the entity association degree between all associated entities and the entity, set an entity association degree threshold, when the entity association degree is higher than the entity association degree threshold, mark the corresponding associated entity as a strongly associated entity, sort all strongly associated entities in descending order of entity association degree value, and mark the sorted sequence as the entity association sequence of the exploration equipment.

[0014] Furthermore, the process for determining whether an exploration equipment exhibits anomalies in exploration signals is as follows: Select an exploration equipment, and sequentially determine the actual data corresponding to each strongly associated entity according to the entity association sequence. For each strongly associated entity whose actual data is determined, obtain and update the exploration signal anomaly index for that exploration equipment. When the exploration signal anomaly index is higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment exhibits an exploration signal anomaly. When the exploration signal anomaly index is not higher than the exploration signal anomaly threshold index, determine the actual data of the next strongly associated entity in the entity association sequence and update the exploration signal anomaly index for that exploration equipment. If, after determining the actual data of all strongly associated entities in the entity association sequence, the updated exploration signal anomaly index is still not higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment does not exhibit an exploration signal anomaly.

[0015] Furthermore, the update steps for the exploration signal anomaly index of the exploration equipment are as follows: Obtain the determined measured deviation ratios among all strongly correlated entities in the entity association sequence of the exploration equipment; sum and average all determined measured deviation ratios to calculate the average measured deviation ratio (BLcg); pair all strongly correlated entities with determined measured deviation ratios into a strong correlated entity combination; when two strongly correlated entities in a strong correlated entity combination are of the same entity type, label the strong correlated entity combination as a similar entity combination; and then obtain the similar deviation coordination coefficient for each similar entity combination. When two strongly correlated entities in a strong correlated entity combination are not of the same entity type, label the strong correlated entity combination as a dissimilar entity combination; and then obtain the dissimilar deviation coordination coefficient for each dissimilar entity combination. Sum and average all similar deviation coordination coefficients to calculate the average similar deviation coordination coefficient. The average coefficient of coordination of all different types of deviations is calculated by summing and averaging all the coefficients of coordination of different types of deviations. Through formula Calculate the exploration signal anomaly index of the exploration equipment. .

[0016] Furthermore, the process of obtaining the similar deviation synergy coefficient of a group of similar related entities is as follows: Select a group of similar related entities, and label the measured deviation ratios of the two strongly related entities in the group as BL1 and BL2, respectively. Calculate the similar deviation synergy coefficient of this group of related entities. .

[0017] Furthermore, the process of obtaining the heterogeneous deviation synergy coefficient of a heterogeneous entity combination is as follows: Select a heterogeneous entity combination, and label the measured deviation ratios of the two strongly related entities in the combination as BL3 and BL4, respectively. Calculate the heterogeneity deviation synergy coefficient of this heterogeneous associated entity combination. ;in, The degree of entity association between BL3 and BL4.

[0018] Furthermore, the calculation process for the measured deviation ratio of a strongly correlated entity is as follows: Select a strongly correlated entity, obtain the entity exploration standard range of the strongly correlated entity, sum and average the two endpoints of the entity exploration standard range to calculate the standard mean, calculate the absolute difference between the two endpoints to obtain the endpoint difference value, and then use the formula... The measured deviation ratio of the strongly correlated entity is calculated.

[0019] Furthermore, when at least one exploration device exhibits an anomaly in its exploration signal, the geological exploration communication index is calculated as follows: The total number of exploration devices exhibiting anomalies is labeled as Nucy. All exploration devices exhibiting anomalies are compared pairwise to obtain the exploration association strength between each pair of compared devices. The average exploration association strength is calculated by summing all exploration association strengths. The geological exploration communication index Fwz is then calculated using Fwz = Nucy * Dep.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] The extraction layer in the knowledge graph of the geological exploration field constructed by this system realizes dynamic quantitative updates of entity association degree. Unlike the traditional fixed association degree setting, the association strength is continuously adjusted with the accumulation of multi-source geological data, so that the association relationship between entities always fits the dynamic changes of geological exploration business, and solves the problem that the traditional static association model cannot adapt to the dynamic evolution of geological data.

[0022] Regularly updating the entity association sequence of each exploration device can promptly filter out core strongly associated entities and eliminate invalid associations, ensuring that anomaly analysis always focuses on entities strongly related to the device's business and avoids identification bias caused by interference from irrelevant entities. At the same time, it eliminates the need to traverse all associated entities for each anomaly analysis, reducing redundant calculations and effectively balancing the dual requirements of "accuracy" and "efficiency" for anomaly identification in geological exploration.

[0023] The number of abnormal devices and the strength of the correlation between devices are quantified into a communication index. Then, the trigger objects with different permissions are matched by the index range to achieve a precise correspondence between "abnormal severity - handling subject permissions". This avoids the waste of high-privilege resources for minor abnormalities and also prevents the handling of major abnormalities from being delayed due to low-privilege responses, making communication triggering more targeted and reasonable. Attached Figure Description

[0024] Figure 1 This is a block diagram illustrating the principle of a knowledge graph-based geological exploration anomaly signal identification and communication triggering system.

[0025] Figure 2 A flowchart for updating the entity association sequence of exploration equipment;

[0026] Figure 3 A flowchart for determining whether there are abnormal exploration signals in exploration equipment. Detailed Implementation

[0027] Reference Figures 1 to 3The geological exploration anomaly signal identification and communication triggering system based on knowledge graph includes a geological exploration knowledge graph construction module, an entity association sequence update module, an exploration anomaly signal identification module, and an anomaly communication triggering module.

[0028] The Geological Exploration Knowledge Graph Construction Module is used to construct a knowledge graph in the field of geological exploration.

[0029] The technical framework for a knowledge graph in the field of geological exploration adopts a six-level progressive architecture: "Data Layer → Extraction Layer → Fusion Layer → Storage Layer → Reasoning Layer → Update Layer." The functions of each layer are as follows: Data Layer: Responsible for the collection and preprocessing of multi-source geological data, outputting standardized data. Extraction Layer: Extracts entities, relationships, and attributes from standardized data to form raw knowledge units. Fusion Layer: Handles knowledge conflicts and redundancy, achieving entity alignment and knowledge completion, outputting high-quality knowledge units. Storage Layer: Uses a graph database to store knowledge units, supporting efficient semantic queries and relational traversal. Reasoning Layer: Expands the knowledge boundary of the graph and uncovers hidden relationships through rule-based reasoning and machine learning reasoning. Update Layer: Establishes an update and review mechanism to ensure the long-term availability of the graph.

[0030] The implementation steps of the data layer are: multi-source geological data acquisition and standardized preprocessing.

[0031] Multi-source geological data acquisition is conducted using the following methods: For geological exploration business scenarios, three types of core data are collected, including structured data (specifically, regional stratigraphic distribution tables (stratigraphic names, densities, and thicknesses), past anomaly signal record tables (signal parameters, anomaly types), and equipment ledgers (equipment models, sampling rates)), semi-structured data (specifically, exploration equipment logs (XML / JSON format, including timestamps and signal amplitudes) and field exploration reports (PDF table fragments)), and unstructured data (specifically, geological expert experience documents (such as "characteristic descriptions of sandstone stratigraphic anomaly signals") and academic literature ("fault and signal correlation studies")).

[0032] Standardized preprocessing methods are as follows: Data cleaning: Structured data: Remove duplicate records (such as duplicate ledgers for the same equipment) and fill in missing values ​​(e.g., fill in a single missing data point with the regional average stratigraphic density). Semi-structured data: Standardize parameter units (e.g., convert "signal amplitude mV" to "V") and filter invalid logs (e.g., empty logs when the equipment is powered off). Unstructured data: Use NLP tools to remove text noise (e.g., meaningless punctuation, repeated sentences), and break long documents into shorter sentences according to the theme of "strata-signal-equipment". Data annotation: Perform domain entity annotation on unstructured text, using a combination of "expert annotation + model pre-annotation": first, use the BERT-BiLSTM-CRF model to pre-annotate entities such as "strata name", "signal parameters", and "equipment model", and then have geological experts review and correct them. For real-time acquired signal data, associate the acquisition device ID with the acquisition location (latitude and longitude) to form standardized data units bound by "data-equipment-location".

[0033] The implementation steps of the extraction layer are: extract knowledge triples (entity-relationship-attribute), and then... Dynamically quantify entity correlation; among which, , , All are weighting coefficients. Since co-occurrence frequency is the fundamental data support for entity association, and the statistical regularity of entity co-occurrence in geological data most directly reflects the essence of the association, it has the highest weight. Semantic similarity, as a supplement when data volume is insufficient, strengthens association judgment through linguistic logic, and has the second highest priority. Domain dependence, as the calibration of data bias by expert knowledge, plays an auxiliary corrective role and has the lowest weight. The three factors work together to balance data objectivity and domain professionalism. The value can be 0.5. The value can be 0.3. The value can be 0.2; Entity co-occurrence count (defined as the number of times an entity co-occurs in multi-source data collected, such as historical reports, device logs, and expert documents). and (Total number of times they occur simultaneously) For entities Total number of occurrences (i.e., the number of times an entity appears in all data) (Total number of times it appears alone or together with other entities) Entity semantic similarity (calculated using natural language processing models such as Word2Vec and BERT) and In terms of semantic similarity, [0,1], the closer the value is to 1, the stronger the correlation between the two entities in terms of linguistic description. Entity domain dependency (the strength of the inherent dependency relationship between two entities, defined by domain experts according to business rules, with a value range of [0,1], and includes core related entity pairs (such as "gravity anomaly signal" and "high-density ore body"). Values ​​close to 1; non-core related entity pairs (such as "meteorological data" and "stratigraphic signal") (Value close to 0).

[0034] The knowledge triples (entity-relationship-attribute) are extracted using the following method: A "multi-model fusion" strategy is adopted, with differentiated extraction methods designed for different data types to ensure the integrity of knowledge units: Entity extraction: Structured data: Entities are directly extracted from fields, such as extracting "sandstone" and "shale" from the "stratum name" field of the "stratum distribution table". Semi-structured / unstructured data: A "BERT-BiLSTM-CRF + domain dictionary" model is used, embedding a geological exploration domain dictionary into the model to improve entity recognition recall. This ultimately forms three core entity libraries: Basic entities: strata (sandstone, shale), equipment (seismographs, gravimeters), signals (vibration signals, gravity signals). Event entities: anomalous events (fault activity, equipment failure), exploration events (field sampling, data acquisition). Attribute entities: parameters (density, frequency), location (exploration area A, exploration line B). Relation Extraction: Based on the "Attention-RCNN Relation Classification Model," the input is "entity pairs + contextual text," and the output is the relationship type between entities, including: membership relationship: "Seismograph S1 - Affiliation - Exploration Team A," acquisition relationship: "Seismograph S1 - Acquisition - Vibration Signal X," correspondence relationship: "Vibration Signal X - Correspondence - Sandstone Strata," and influence relationship: "Fault Activity - Influence - Signal Amplitude." For structured data, relationships are directly determined based on inter-table associations (e.g., association between "Equipment Ledger ID" and "Log Equipment ID"); for unstructured data, relationships are determined through contextual semantics (e.g., "Seismograph acquired gravity signals in exploration area A"). Attribute Extraction: Using a "rule matching + attribute value extraction model," key domain attributes are extracted for each entity: Stratigraphic entities: density (ρ, unit g / cm³), thickness (h, unit m), distribution depth (d, unit m). Signal entities: frequency (f, unit Hz), amplitude (A, unit V), duration (t, unit s). Equipment Entities: Sampling Rate (r, in Hz), Operating Voltage (U, in V), Detection Accuracy (ε, in %). Example: Extract "Sandstone - Density = 2.6 g / cm³" and "Sandstone - Thickness = 50 m" from the text "Sandstone strata density is approximately 2.6 g / cm³, thickness 50 m".

[0035] The implementation steps of the fusion layer are: knowledge conflict handling and entity alignment.

[0036] The specific methods for handling knowledge conflicts are as follows:

[0037] Attribute Conflict: Set data priority: Expert-annotated data > Real-time exploration data > Historical database data. Example: If the expert-annotated "shale density 2.5g / cm³" conflicts with the historical data "shale density 2.3g / cm³", the expert-annotated data shall prevail. If the priorities are the same (e.g., two different expert-annotated "sandstone thickness" values ​​are 60m and 70m respectively), the average value shall be taken as the final attribute value (65m), and the original data source shall be recorded in the map.

[0038] Relationship Conflict: Employing "Multi-Model Voting": Attention-RCNN, BERT-RE, and PCNN are used to determine the relationship of the same entity pair. The final relationship is determined when "more than 2 / 3 of the models agree". Example: If Attention-RCNN determines "signal Y - corresponds to - sandstone", BERT-RE determines "signal Y - corresponds to - shale", and PCNN determines "signal Y - corresponds to - sandstone", then the final relationship is "signal Y - corresponds to - sandstone".

[0039] Entity alignment is performed using the following methods: String matching: Fuzzy matching of entity names (e.g., the matching degree of "Seismograph S1" and "Equipment S1" is ≥80%) to initially screen for suspected duplicate entities; Attribute matching: Comparing the core attributes of suspected entities (e.g., the "sampling rate" and "model" of the equipment). If the attribute overlap is ≥90%, they are determined to be the same entity; Association matching: Checking whether the associated entities of suspected entities are consistent (e.g., "Seismograph S1" is associated with "Signal X", and "Equipment S1" is also associated with "Signal X"). If the association overlap is ≥85%, alignment is confirmed and they are merged into the same entity.

[0040] The implementation steps of the storage layer are: graph database selection (using Neo4j graph database as the storage medium) and storage structure design.

[0041] The storage structure design follows these steps: Each entity corresponds to a node, with node attributes including "Entity ID," "Name," "Type," "Core Attributes," "Data Source," and "Creation Time." Example: Stratigraphy node (ID: STR001, Name: Sandstone, Type: Basic Entity - Stratigraphy, Density: 2.6g / cm³, Source: Expert Annotation, Creation Time: 2025-01-10). Each relationship corresponds to an edge, with edge attributes including "Relationship ID," "Relationship Type," and "Association Degree." Index design: Indexes are created for frequently queried fields, including "Entity ID," "Entity Type," and "Signal Parameters (Frequency, Amplitude)."

[0042] The implementation steps of the reasoning layer are to formulate multiple reasoning rules in the field of geological exploration, and deduce new associations from existing knowledge (not elaborated here, but only illustrated by example: Rule 1: If “Equipment A - Acquisition - Signal B” and “Signal B - Corresponding - Formation C”, then “Equipment A - Detection - Formation C” is deduced).

[0043] The update layer is used for periodic batch updates, including cleaning up invalid knowledge (such as equipment ledgers that have not been updated for more than 5 years).

[0044] The entity association sequence update module updates the entity association sequence of each exploration device periodically based on the knowledge graph of the geological exploration field. (The time interval corresponding to the periodicity is set and adjusted according to the geological risk of the geological exploration area. For example, if the geological exploration area is in an active seismic zone, the time interval can be set to be shorter; if the geological exploration area is in a general mineral exploration area, the time interval can be set to be longer. This will not be elaborated here.)

[0045] The exploration anomaly signal identification module determines whether there are exploration signal anomalies in each exploration device. If there are no exploration signal anomalies in all exploration devices, communication does not need to be triggered. If at least one exploration device has an exploration signal anomaly, the geological exploration communication index is calculated.

[0046] When at least one exploration device exhibits an anomaly in its exploration signal, the geological exploration communication index is calculated as follows: The total number of exploration devices exhibiting anomalies is labeled as Nucy. All exploration devices exhibiting anomalies are compared pairwise to obtain the exploration association strength between each pair of compared devices. The average exploration association strength is calculated by summing all exploration association strengths and then calculating the average exploration association strength Dep. The geological exploration communication index Fwz is calculated using Fwz=Nucy*Dep.

[0047] The process for obtaining the exploration correlation strength between the two compared exploration devices is as follows:

[0048] Step 1: Identify all "relationship edges" in the association path between the two compared exploration devices. In the knowledge graph, the association path between devices consists of a series of "entity-relationship-entity" edges. For example, the two compared exploration devices are device A and device B, and the path "device A-acquisition-signal X-correspondence-stratum Y-association-device B" contains 3 edges: Edge 1: device A → (acquisition, entity association degree R1) → signal X; Edge 2: signal X → (correspondence, entity association degree R2) → stratum Y; Edge 3: stratum Y → (association, entity association degree R3) → device B.

[0049] Step 2: Extract the "entity correlation degree" of each edge; taking the above 3 edges as an example, the following need to be extracted from the graph: R1: entity correlation degree between device A and signal X (e.g., 0.8); R2: entity correlation degree between signal X and formation Y (e.g., 0.9); R3: entity correlation degree between formation Y and device B (e.g., 0.7).

[0050] Step 3: Take the "minimum" of the correlation between all entities; in the example, min(0.8,0.9,0.7)=0.7, so the exploration correlation strength between device A and device B is 0.7.

[0051] The update process of an entity association sequence for an exploration device is as follows: Select an exploration device, determine its entity position in the geological exploration knowledge graph, mark all entities in the geological exploration knowledge graph that are associated with the device but do not belong to attribute information as associated entities (excluding entities that belong to attribute information (such as entity ID, region number, etc.), ensuring that associated entities are independent objects that can generate business associations with the target entity), obtain the entity association degree between all associated entities and the target entity, set the entity association degree threshold (the entity association degree threshold is set according to the pattern of historical association data, which will not be elaborated here), when the entity association degree is higher than the entity association degree threshold, mark the corresponding associated entity as a strongly associated entity (if it is not higher, no marking is required), sort all strongly associated entities in descending order of entity association degree value, and mark the sorted sequence as the entity association sequence of the exploration device.

[0052] The procedure for determining whether there are abnormal exploration signals in exploration equipment is as follows: Select an exploration device, and determine the actual data corresponding to each strongly correlated entity according to the entity association sequence (the entities in the entity association sequence (signal type, stratum type, equipment status type, business type), although different in type, all have actual data - these data are all measured records, as shown in the following examples: Signal type entities (such as vibration signals, microseismic signals), corresponding actual data: measured time domain / frequency domain parameters; Stratum type entities (such as sandstone strata, fault strata), corresponding actual data: field survey measured data (such as density 2.6g / cm³, thickness 50m, distribution depth 200-300m); Equipment status type (such as abnormal sampling rate, normal voltage), corresponding actual data... Actual data: measured data from equipment sensors (e.g., sampling rate of 950Hz during abnormal conditions, voltage of 220V during normal conditions). For each strongly correlated entity, the actual data is determined, and the exploration signal anomaly index of the exploration equipment is updated. When the exploration signal anomaly index is higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment has an exploration signal anomaly. When the exploration signal anomaly index is not higher than the exploration signal anomaly threshold index, the actual data of the next strongly correlated entity in the entity correlation sequence is determined, and the exploration signal anomaly index of the exploration equipment is updated. If, after determining the actual data of all strongly correlated entities in the entity correlation sequence, the updated exploration signal anomaly index is still not higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment does not have an exploration signal anomaly.

[0053] The update steps for the exploration signal anomaly index of exploration equipment are as follows: Obtain the determined measured deviation ratios among all strongly correlated entities in the entity association sequence of the exploration equipment. Sum and average all determined measured deviation ratios to calculate the average measured deviation ratio (BLcg). Pairwise match all strongly correlated entities with determined measured deviation ratios into a strong correlated entity combination. When two strongly correlated entities in a strong correlated entity combination are of the same entity type, label the strong correlated entity combination as a similar entity combination, and then obtain the similar deviation coordination coefficient for each similar entity combination. When two strongly correlated entities in a strong correlated entity combination are not of the same entity type, label the strong correlated entity combination as a dissimilar entity combination, and then obtain the dissimilar deviation coordination coefficient for each dissimilar entity combination. Sum and average all similar deviation coordination coefficients to calculate the average similar deviation coordination coefficient. The average coefficient of coordination of all different types of deviations is calculated by summing and averaging all the coefficients of coordination of different types of deviations. Through formula Calculate the exploration signal anomaly index of the exploration equipment. .

[0054] The process of obtaining the similar deviation synergy coefficient of a group of similar related entities: Select a group of similar related entities, and label the measured deviation ratios of the two strongly related entities in the group as BL1 and BL2, respectively. Calculate the similar deviation synergy coefficient of this group of related entities. .

[0055] The process of obtaining the heterogeneous deviation synergy coefficient of a heterogeneous entity combination: Select a heterogeneous entity combination, and label the measured deviation ratios of the two strongly related entities in the combination as BL3 and BL4, respectively. Calculate the heterogeneity deviation synergy coefficient of this heterogeneous associated entity combination. ;in, This represents the entity association degree between two strongly related entities in a heterogeneous entity combination.

[0056] The calculation process for the measured deviation ratio of a strongly correlated entity is as follows: Select a strongly correlated entity and obtain its entity exploration standard range (the entity exploration standard range is a pre-stored "normal attribute interval" in the geological exploration knowledge graph, for example, the entity exploration standard range for sandstone density is 2.4-2.8 g / cm³). Calculate the average of the two endpoints of the entity exploration standard range (the endpoints are the two extreme values ​​of the entity exploration standard range), and obtain the standard mean. Calculate the absolute difference between the two endpoints to obtain the endpoint difference value, and then use the formula... The measured deviation ratio of the strongly correlated entity is calculated.

[0057] The abnormal communication triggering module determines the communication triggering object based on the geological exploration communication index and sends abnormal exploration signals to the corresponding communication triggering object.

[0058] The communication triggering objects are determined based on the geological exploration communication index, as follows: Each geological exploration communication index Fwz is set to correspond to a different communication triggering object. The range of the geological exploration communication index is (0, Fw1], (Fw1, Fw2], ..., (FwZ-1, FwZ], and the communication triggering objects include communication triggering object 1, communication triggering object 2, ..., communication triggering object Z-1, and communication triggering object Z. When the geological exploration communication index is in (0, Fw1], it corresponds to communication triggering object 1. The responsibility authority of communication triggering object 1 is lower than that of communication triggering object 2, and so on. For example, communication triggering object 1 is the regional exploration station technician, communication triggering object 2 is the regional exploration station head, and communication triggering object 3 is the geological emergency command center.

[0059] The above formulas are all dimensionless calculations, and the preset parameters in the formulas should be set by those skilled in the art according to the actual situation.

[0060] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0061] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0062] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0063] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0064] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0065] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A knowledge graph-based geological exploration anomaly signal identification and communication triggering system, characterized in that, include: The geological exploration knowledge graph construction module is used to construct a knowledge graph in the field of geological exploration. The entity association sequence update module updates the entity association sequence of each exploration device regularly based on the knowledge graph in the geological exploration field. The exploration anomaly signal identification module is used to determine whether there are exploration signal anomalies in each exploration device. If there are no exploration signal anomalies in all exploration devices, communication does not need to be triggered. If at least one exploration device has an exploration signal anomaly, the geological exploration communication index is calculated. The abnormal communication triggering module determines the communication triggering object based on the geological exploration communication index and sends abnormal exploration signals to the corresponding communication triggering object.

2. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 1, characterized in that, The technical framework for knowledge graphs in the field of geological exploration adopts a six-level progressive architecture: "Data Layer → Extraction Layer → Fusion Layer → Storage Layer → Reasoning Layer → Update Layer." The extraction layer's implementation steps include: extracting knowledge triples, and then... Dynamically quantify entity correlation; among which, , , All are weighting coefficients. ; The number of times an entity co-occurs; For entities Total number of occurrences; For entity semantic similarity; Dependence on the physical domain.

3. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 1, characterized in that, The update process of an entity association sequence for an exploration device is as follows: Select an exploration device, determine the entity position of the exploration device in the geological exploration domain knowledge graph, mark all entities in the geological exploration domain knowledge graph that are associated with the entity but are not attribute information as associated entities, obtain the entity association degree between all associated entities and the entity, set an entity association degree threshold, when the entity association degree is higher than the entity association degree threshold, mark the corresponding associated entity as a strongly associated entity, sort all strongly associated entities in descending order of entity association degree value, and mark the sorted sequence as the entity association sequence of the exploration device.

4. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 1, characterized in that, The process for determining whether an exploration equipment has an anomaly in its exploration signal is as follows: Select an exploration equipment, and determine the actual data corresponding to each strongly associated entity according to the entity association sequence. For each strongly associated entity whose actual data is determined, obtain and update the exploration signal anomaly index of the exploration equipment. When the exploration signal anomaly index is higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment has an exploration signal anomaly. When the exploration signal anomaly index is not higher than the exploration signal anomaly threshold index, determine the actual data of the next strongly associated entity in the entity association sequence and update the exploration signal anomaly index of the exploration equipment. If the actual data of all strongly associated entities in the entity association sequence are determined, and the updated exploration signal anomaly index is still not higher than the exploration signal anomaly threshold index, it indicates that the exploration equipment does not have an exploration signal anomaly.

5. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 4, characterized in that, The update steps for the exploration signal anomaly index of exploration equipment are as follows: Obtain the determined measured deviation ratios among all strongly correlated entities in the entity association sequence of the exploration equipment. Sum and average all determined measured deviation ratios to calculate the average measured deviation ratio (BLcg). Pairwise match all strongly correlated entities with determined measured deviation ratios into a strong correlated entity combination. When two strongly correlated entities in a strong correlated entity combination are of the same entity type, label the strong correlated entity combination as a similar entity combination, and then obtain the similar deviation coordination coefficient for each similar entity combination. When two strongly correlated entities in a strong correlated entity combination are not of the same entity type, label the strong correlated entity combination as a dissimilar entity combination, and then obtain the dissimilar deviation coordination coefficient for each dissimilar entity combination. Sum and average all similar deviation coordination coefficients to calculate the average similar deviation coordination coefficient. The average coefficient of coordination of all different types of deviations is calculated by summing and averaging all the coefficients of coordination of different types of deviations. Through formula Calculate the exploration signal anomaly index of the exploration equipment. .

6. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 5, characterized in that, The process of obtaining the similar deviation synergy coefficient of a group of similar related entities: Select a group of similar related entities, and label the measured deviation ratios of the two strongly related entities in the group as BL1 and BL2, respectively. Calculate the similar deviation synergy coefficient of this group of related entities. .

7. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 5, characterized in that, The process of obtaining the heterogeneous deviation synergy coefficient of a heterogeneous entity combination: Select a heterogeneous entity combination, and label the measured deviation ratios of the two strongly related entities in the combination as BL3 and BL4, respectively. Calculate the heterogeneity deviation synergy coefficient of this heterogeneous associated entity combination. ;in, This represents the entity association degree between two strongly related entities in a heterogeneous entity combination.

8. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 5, characterized in that, The calculation process for the measured deviation ratio of a strongly correlated entity is as follows: Select a strongly correlated entity, obtain the entity's standard exploration range, sum and average the two endpoints of the standard exploration range to obtain the standard mean, calculate the absolute difference between the two endpoints to obtain the endpoint difference value, and then use the formula... The measured deviation ratio of the strongly correlated entity is calculated.

9. The knowledge graph-based geological exploration anomaly signal identification and communication triggering system according to claim 1, characterized in that, When at least one exploration device exhibits an anomaly in its exploration signal, the geological exploration communication index is calculated as follows: The total number of exploration devices exhibiting anomalies is labeled as Nucy. All exploration devices exhibiting anomalies are compared pairwise to obtain the exploration association strength between each pair of compared devices. The average exploration association strength is calculated by summing all exploration association strengths and then calculating the average exploration association strength Dep. The geological exploration communication index Fwz is calculated using Fwz=Nucy*Dep.

Images