Intelligent matching diagnosis method and system based on standard-device double knowledge bases

Abstract

The application relates to an intelligent matching diagnosis method and system based on a standard-equipment double knowledge base, which comprises the following steps: constructing a standard-equipment double knowledge base comprising a standard knowledge base and an equipment knowledge base, and combining a Neo4j graph database to build a knowledge graph; collecting equipment operation data in real time through an edge computing node arranged on a device control cabinet, pre-processing the equipment operation data, and importing the obtained equipment operation data into a lightweight rule engine; incrementally updating the knowledge graph according to the standard change content detected in real time, and selecting affected equipment from the updated knowledge graph according to the matching result for processing. Through the standard-equipment associated knowledge graph, the dynamic association of a standard database set and an equipment database is realized, compliance diagnosis is upgraded from "manual comparison" to "system automatic matching", and the efficiency is improved; the edge computing node and the lightweight rule engine are arranged, the operation data and the standard are matched at a minute level, and potential compliance risks can be warned in advance.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence knowledge base construction in the power generation sector, and in particular to an intelligent matching and diagnostic system based on a standard-equipment dual knowledge base. Background Technology

[0002] Currently, technical supervision in the power generation sector mainly relies on manual verification, which has three major pain points: First, technical supervision standards (such as equipment vibration limits and energy consumption index regulations) are statically stored in document form, requiring manual retrieval and matching of corresponding equipment, which is inefficient and prone to errors due to misunderstandings of standard clauses; second, the comparison of equipment operating data with standard thresholds relies on manual analysis after the fact, which cannot detect compliance risks in real time and is prone to causing equipment failures or regulatory penalties; third, when technical standards are updated (such as adding new energy consumption standards or adjusting safety limits), the impact of standard changes on existing equipment needs to be assessed manually, which is time-consuming, labor-intensive, and difficult to comprehensively cover all related equipment.

[0003] Although existing technologies include standard query systems or equipment monitoring systems, none of them have achieved deep correlation and dynamic matching between standards and equipment, which cannot meet the needs of power generation companies to transform their technical supervision work towards intelligence and precision. Summary of the Invention

[0004] This application provides an intelligent matching and diagnostic method and system based on a standard-device dual knowledge base, which at least solves the problem that related technologies cannot meet the needs of enterprises for efficient compliance management.

[0005] In a first aspect, embodiments of this application provide an intelligent matching and diagnostic method based on a standard-device dual knowledge base, including: Construct a standard-device dual knowledge base that includes a standard knowledge base and a device knowledge base, and build a knowledge graph by combining the Neo4j graph database; By collecting equipment operation data in real time through edge computing nodes deployed on the equipment control cabinet, the equipment operation data is preprocessed and then imported into a lightweight rule engine for matching to obtain matching results. The knowledge graph is incrementally updated based on the standard changes detected in real time. Affected devices are selected from the updated knowledge graph based on the matching results, and then processed according to the type of the affected devices.

[0006] Preferably, the construction of a standard-device dual knowledge base, including a standard knowledge base and a device knowledge base, and the building of a knowledge graph using the Neo4j graph database, includes: Obtain national supervision standards or industry technical supervision standards, construct the standard knowledge base based on the national supervision standards, construct the equipment knowledge base based on the industry technical supervision standards, and associate the standard knowledge base and the equipment knowledge base to obtain the standard-equipment dual knowledge base; Based on the equipment knowledge base, a hierarchical relationship model containing equipment, components, rated parameters, and operation and maintenance status is constructed to build a four-element relationship model, and the multi-dimensional association rules in the four-element relationship model are determined. The knowledge graph is constructed based on the Neo4j graph database. The knowledge graph is then logically validated according to the multi-dimensional association rules. After the validation is completed, a visualized knowledge graph is generated.

[0007] Preferably, the step of constructing the knowledge graph based on the Neo4j graph database, performing logical verification on the knowledge graph according to the multi-dimensional association rules, and generating a visualized knowledge graph after verification includes: The knowledge graph is constructed based on the Neo4j graph database, and the knowledge graph includes a basic layer, a relational layer, and a derivative layer. The multi-dimensional association rules are imported into the knowledge graph reasoning engine, and the knowledge graph reasoning engine is used to perform consistency verification, integrity verification and rationality verification in sequence. After the verification is completed, the visualized knowledge graph is generated.

[0008] Preferably, the step of collecting equipment operation data in real time through edge computing nodes deployed on the equipment control cabinet, preprocessing the equipment operation data, and importing the obtained equipment operation data into a lightweight rule engine for matching to obtain matching results includes: The edge computing nodes are deployed in the equipment control cabinet. Based on the edge computing nodes, the equipment operation data is collected in real time. The equipment operation data is then preprocessed, including wavelet denoising, outlier removal, and data standardization, to obtain the processed equipment operation data. The lightweight rule engine is customized based on Drools rules. The device operation data is imported into the customized lightweight rule engine to perform matching priority management and obtain the matching result.

[0009] Preferably, the step of customizing the lightweight rule engine based on Drools rules, importing the device operation data into the customized lightweight rule engine to perform matching priority management, and obtaining the matching result includes: Based on the Drools rules, redundant modules in the lightweight rule engine are removed to obtain the customized lightweight rule engine; The lightweight rule engine performs multi-dimensional matching, including trend matching and device group matching. For scenarios where standard clauses are vaguely worded, a fuzzy membership function is constructed by combining historical equipment operation data to achieve quantitative matching of qualitative standards, thus obtaining the matching result.

[0010] Preferably, the step of incrementally updating the knowledge graph based on the standard changes detected in real time, selecting affected devices from the updated knowledge graph based on the matching results, and processing the affected devices according to their types includes: Based on the standard update monitoring trigger, when the standard is updated, the updated standard document is downloaded, the changes are identified, and new, revised and obsolete clauses are marked. A device-parameter-cost-risk assessment model is constructed, and Monte Carlo simulation is used to predict the probability of the device meeting the standard at the effective time point. Based on the probability of meeting the standard, affected devices are screened out, and the affected devices are processed according to different types.

[0011] Preferably, the constructed equipment-parameter-cost-risk assessment model uses Monte Carlo simulation to predict the probability of equipment meeting the standards at the effective time point, filters out affected equipment based on the compliance probability, and processes the affected equipment according to different types, including: Construct the equipment-parameter-cost-risk assessment model, input equipment parameters into the equipment-parameter-cost-risk assessment model, and use the Monte Carlo simulation to predict the probability of the equipment meeting the standards at the effective time point; The system monitors whether the equipment parameters meet the standards through real-time diagnostics. If they do, a change execution acceptance report is generated; if they do not, the reasons are analyzed. All change-related documents are automatically archived in the enterprise document management system, forming a traceable compliance management file.

[0012] Secondly, embodiments of this application provide an intelligent matching and diagnostic system based on a standard-device dual knowledge base, comprising: The knowledge fusion module is used to build a standard-device dual knowledge base, which includes a standard knowledge base and a device knowledge base, and to build a knowledge graph in conjunction with the Neo4j graph database. The real-time diagnostic module is used to collect equipment operation data in real time through edge computing nodes deployed on the equipment control cabinet, preprocess the equipment operation data, import the obtained equipment operation data into a lightweight rule engine for matching, and obtain matching results. The dynamic update module is used to incrementally update the knowledge graph based on the standard changes detected in real time, select the affected devices from the updated knowledge graph based on the matching results, and process the affected devices according to their types.

[0013] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the intelligent matching and diagnostic method based on a standard-device dual knowledge base as described in the first aspect above.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent matching and diagnostic method based on a standard-device dual knowledge base as described in the first aspect above.

[0015] The intelligent matching and diagnostic system based on a standard-device dual knowledge base provided in this application has at least the following technical effects: By using a knowledge graph that links standards and devices, dynamic association between the two is achieved, upgrading compliance diagnosis from "manual comparison" to "automatic system matching," thus improving efficiency. Deploying edge computing nodes and a lightweight rule engine enables minute-level matching of operational data with standards, providing early warnings of potential compliance risks and avoiding regulatory penalties.

[0016] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of an intelligent matching and diagnostic method based on a dual knowledge base of standards and devices, provided by relevant technologies. Figure 2 This is a flowchart of step S20 provided by relevant technology; Figure 3 This is a flowchart of step S30 provided by relevant technology; Figure 4 This is a structural block diagram of an intelligent matching and diagnostic system based on a standard-device dual knowledge base, provided by relevant technologies; Figure 5 It is a structural diagram of an electronic device provided based on relevant technologies. Detailed Implementation

[0018] Firstly, embodiments of this application provide an intelligent matching and diagnostic method based on a standard-device dual knowledge base. Figure 1 This is a flowchart of the method, such as... Figure 1 As shown, it includes: Step S10: Construct a standard-device dual knowledge base, which includes a standard knowledge base and a device knowledge base, and build a knowledge graph using the Neo4j graph database.

[0019] Step S20: Real-time collection of equipment operation data is performed by edge computing nodes deployed on the equipment control cabinet. The equipment operation data is preprocessed and then imported into a lightweight rule engine for matching to obtain matching results.

[0020] Step S30: Incrementally update the knowledge graph based on the standard changes detected in real time, select the affected devices from the updated knowledge graph based on the matching results, and process them according to the type of the affected devices.

[0021] By establishing a knowledge graph linking standards and devices, dynamic correlation between the two is achieved, upgrading compliance diagnostics from manual comparison to automatic system matching, thus improving efficiency. Deploying edge computing nodes and a lightweight rule engine enables minute-level matching of operational data with standards, providing early warnings of potential compliance risks and avoiding regulatory penalties.

[0022] Specifically, step S10 includes: (1) Obtain national supervision standards or industry technical supervision standards, build a standard knowledge base based on national supervision standards, build an equipment knowledge base based on industry technical supervision standards, and link the standard knowledge base and equipment knowledge base to obtain a standard-equipment dual knowledge base.

[0023] Standard knowledge base construction: Collect national / industry technical supervision standards (such as GB / T 11348 "Measurement and Evaluation of Radial Vibration of Rotating Machinery Shafts" and DL / T 5047 "Technical Specifications for Construction and Acceptance of Power Plants"), first establish a standard classification system (divided into 4 major categories and 23 subcategories according to "safety / environmental protection / equipment performance / energy consumption"), then use OCR recognition (using the Tesseract 5.0 engine, with an accuracy rate ≥98.5%) and NLP semantic parsing (based on a BERT pre-trained model, with an entity extraction F1 value ≥0.92 after fine-tuning) to extract entities (such as "vibration limits", "temperature thresholds", and "applicable equipment types") and attributes (such as limit values, units, applicable operating conditions, effective date, and repeal conditions) from the standard clauses; manage the extracted structured data for versions, add timestamps and version numbers to each standard clause, and build a standard database that supports historical backtracking.

[0024] Equipment knowledge base construction: Collect enterprise equipment ledger data (equipment model, rated parameters, component composition, installation location, commissioning date), 3D model data (STEP format, accuracy ≤0.1mm), and operation and maintenance history data (fault records, maintenance records, modification records). Access the equipment factory technical manual (PDF / Word format) through the industrial IoT platform. Use entity alignment algorithm (based on Jaccard similarity and cosine similarity fusion strategy) to associate multi-source equipment data and extract the four-level hierarchical relationship of "equipment-component-rated parameters-operation and maintenance status". Standardize equipment parameters, such as converting "speed 3000 rpm" to "3000 r / min", to form a structured equipment dataset. The dataset supports incremental import (maximum import volume of ≥1000 equipment data in a single import).

[0025] (2) Based on the equipment knowledge base, a hierarchical relationship model containing equipment, components, rated parameters and operation and maintenance status is constructed to build a four-element relationship model and determine the multi-dimensional association rules in the four-element relationship model.

[0026] (3) Construct a knowledge graph based on the Neo4j graph database, perform logical verification on the knowledge graph according to multi-dimensional association rules, and generate a visualized knowledge graph after verification.

[0027] The process of generating a knowledge graph is as follows: A knowledge graph is constructed based on the Neo4j graph database. The knowledge graph includes a basic layer, an association layer, and a derivative layer. Multi-dimensional association rules are imported into the knowledge graph reasoning engine, which then performs consistency checks, integrity checks, and rationality checks in sequence. After the checks are completed, a visualized knowledge graph is generated.

[0028] Entity Association Rule Definition: Establish a four-element relationship model of "standard clause - equipment component - operating parameter - association weight" and define multi-dimensional association rules: Basic association rules: For example, “Standard Clause GB / T 11348.1-2009 Clause 5.2” is associated with the “Steam Turbine - High Pressure Rotor” component, with the corresponding operating parameter “Radial Vibration Displacement”, the threshold “≤0.05mm (speed 3000r / min)”, and the association weight is set to 0.9 (weight range 0-1, the higher the value, the higher the association priority). Operating condition adaptation rules: For equipment with variable operating conditions (such as peak-shaving units), define the mapping relationship between "operating condition parameters and standard thresholds", such as "when the turbine load is <50% of the rated power, the vibration limit is relaxed to ≤0.08mm"; Equipment status association rules: The association logic is adjusted based on the equipment operation and maintenance status, such as "when equipment has been in operation for more than 10 years and has accumulated more than 500 start-stop cycles, the standard threshold is tightened by 10%". The entity association logic is defined using the SPARQL query language (which supports custom function extensions). A production rule engine (such as CLIPS) is introduced to build the rule base. The rule base supports visual configuration (editing rule conditions and conclusions through a drag-and-drop interface) to ensure accurate and dynamic mapping between standard clauses and equipment parameters. New association rules undergo syntax validation and conflict detection (detection efficiency ≤ 100 rules / second) to avoid rule contradictions.

[0029] Knowledge Graph Generation and Verification: A knowledge graph is constructed based on the Neo4j graph database (using a cluster deployment mode with ≥3 nodes and support for horizontal scaling), employing a three-layer architecture: ① Base Layer: Contains basic nodes and attributes such as standard entities, equipment entities, parameter entities, and operating condition entities; ② Association Layer: Stores relationships and weights such as "clause-component," "component-parameter," and "parameter-operating condition"; ③ Derivative Layer: Intermediate nodes generated based on reasoning (e.g., "high-risk compliance clause set" and "similar equipment parameter comparison set"). When importing structured data into the graph database, a batch import tool (neo4j-adminimport) is used, with an import rate ≥100,000 records / minute.

[0030] Multi-dimensional logical verification is performed through a knowledge graph reasoning engine (combining Pellet and custom rule reasoning): Consistency verification: Checks whether the thresholds of the same equipment component conflict in different standards, and automatically corrects them according to the principles of "newest published standard first", "industry standard first, stricter than national standard first", and "special working condition standard first, general standard first"; Integrity verification: Identifies missing relationships (such as a certain equipment component not matching any standard clause) and generates a "relationship missing warning list"; Reasonableness verification: Determines whether the parameter thresholds are within a reasonable range through statistical analysis (such as temperature thresholds cannot be negative). After verification, a visual knowledge graph is generated (supporting switching between Force-Directed and Hierarchical layouts), providing node query (response time ≤ 500ms), relationship tracing (supporting multi-level path display), and path analysis functions; at the same time, a knowledge graph index (based on the Lucene engine) is built to improve the efficiency of complex queries (multi-condition combined query response time ≤ 2s). In addition, a knowledge enhancement mechanism is introduced, supplementing entity attributes and relationships by crawling external data such as industry technical forums and equipment manufacturer technical white papers, and automatically updating the enhanced data once a quarter.

[0031] Step S20 for performing edge computing and lightweight matching is as follows Figure 2 As shown, it includes: Step S21: Deploy edge computing nodes in the equipment control cabinet, collect equipment operation data in real time based on the edge computing nodes, and perform preprocessing on the equipment operation data, including wavelet denoising, outlier removal and data standardization, to obtain processed equipment operation data.

[0032] Step S22: Customize the lightweight rule engine based on Drools rules, match the customized lightweight rule engine, and enable the matched lightweight rule engine to perform matching priority management in the matching results to obtain the matching results.

[0033] Step S22 specifically includes: (1) Based on the Drools rules, the redundant modules in the lightweight rule engine are removed to obtain the customized lightweight rule engine; (2) Perform multi-dimensional matching, including trend matching and device group matching, on the lightweight rule engine; (3) For scenarios where the standard clauses are vague, a fuzzy membership function is constructed by combining the historical operating data of the equipment to achieve quantitative matching of the qualitative standard and obtain the matching result.

[0034] Edge node deployment and data acquisition: Edge computing nodes (using NVIDIA Jetson Nano embedded chips) are deployed in the equipment control cabinet to acquire real-time operating data of the equipment via RS485 / Modbus protocol. The acquired parameters include turbine vibration frequency (sampling frequency 100Hz), boiler outlet temperature (accuracy ±1℃), generator stator current (accuracy 0.5 grade), etc. The acquired data is preprocessed, including wavelet denoising (db3 wavelet decomposition), outlier removal (3σ criterion), and data standardization (mapping to the [0,1] interval) to ensure data quality.

[0035] Edge node deployment and data acquisition: Hardware Deployment: A "primary and backup dual-node" redundancy design is adopted. The primary edge node (NVIDIA Jetson Nano embedded chip, computing power 21 TOPS) and the backup node (Raspberry Pi 4B, computing power 1.5 TOPS) are deployed in the equipment control cabinet. The nodes monitor their status through a heartbeat mechanism (1 second interval). If the primary node fails, it will automatically switch to the backup node within 100ms. It is equipped with an industrial-grade power supply module (input DC 24V±10%, output ripple ≤50mV) and a surge protection module (IEC 61000-4-5 Class 4) to adapt to the harsh industrial environment.

[0036] Multi-protocol acquisition: Supports mainstream industrial protocols such as RS485 / Modbus, Profinet, and EtherCAT, and can connect to devices from different manufacturers (such as Siemens S7-400 PLC and Schneider M340 PLC) through a protocol conversion gateway; for older devices without digital interfaces, an analog signal acquisition module (16-bit accuracy, input range 0-10V / 4-20mA) is deployed to convert the signal into a digital signal.

[0037] Dynamic acquisition strategy: The acquisition frequency is automatically adjusted based on the equipment's operating conditions. For example, the acquisition frequency is set to 10Hz when the equipment is operating normally. When the load fluctuation is greater than ±10% of the rated value or the parameter is close to the threshold, high-frequency acquisition (100Hz) is triggered. Edge caching technology (local cache capacity ≥16GB) is adopted to cache data when the network is disconnected and automatically retransmit it after the network is restored to ensure that the data is not lost.

[0038] Enhanced data preprocessing: In addition to wavelet denoising and outlier removal, new trend smoothing processing (using moving average filtering with a window size of 5-10 sampling points) and data alignment correction (based on timestamp synchronization with a synchronization accuracy of ≤1ms) are added. For nonlinear parameters (such as temperature and pressure), piecewise linear interpolation is used to complete missing data. The data validity rate after preprocessing is ≥99.8%.

[0039] Lightweight rule engine matching: Rule engine optimization: Based on the Drools rule engine, it has been trimmed and customized, removing redundant modules (such as complex event processing modules), and adopting rule compilation caching (compiling rules into bytecode and caching them in memory). The startup time is ≤3s and the memory usage is ≤512MB. It supports hot rule updates (no need to restart the engine, update response time is ≤5s), meeting the real-time rule adjustment needs after standard changes.

[0040] Multi-dimensional matching mechanism: In addition to single-parameter and multi-parameter correlation matching, the following new features are added: a) Trend matching: By analyzing the slope of parameter changes through linear regression (e.g., the rate of increase of vibration displacement within 5 minutes > 0.02 mm / min), potential risks of exceeding standards can be identified in advance; b) Equipment group matching: By comparing the same parameters of similar equipment (e.g., #1 and #2 steam turbines), when the parameter of a certain equipment deviates from the group average by more than 20%, an "abnormal deviation of equipment" warning is triggered; c) Fuzzy matching: For scenarios where the standard clauses are vaguely described (e.g., "vibration values ​​should be controlled within a reasonable range"), a fuzzy membership function is constructed by combining the historical operating data of the equipment (the distribution of normal operating parameters in the past 3 months) to achieve quantitative matching of qualitative standards.

[0041] Matching priority management: Establish a four-level matching priority (urgent > high > medium > low). Urgent priority (such as safety valve opening pressure exceeding the standard) will occupy computing resources first, with a matching response time ≤ 50ms; high priority (such as critical equipment temperature exceeding the standard) will have a response time ≤ 100ms, ensuring that high-risk compliance issues are handled first.

[0042] Machine learning-assisted optimization: A lightweight random forest model (training sample size ≥ 100,000) is introduced to learn from historical matching results and automatically optimize rule parameters (such as the weight coefficients of multi-parameter association matching). The model iterates automatically every 7 days, and the matching accuracy improves with the accumulation of data (target ≥ 98%).

[0043] Step S30, used to perform incremental updates to the knowledge graph and process affected devices, such as Figure 3 As shown, it includes: Step S31: When the standard update monitoring trigger detects a standard update, download the updated standard document, identify the changes, and mark the added, revised, and abolished clauses.

[0044] Step S32: Construct an equipment-parameter-cost-risk assessment model, use Monte Carlo simulation to predict the probability of the equipment meeting the standards at the effective time point, screen out the affected equipment based on the probability of meeting the standards, and handle the affected equipment according to different types.

[0045] Step S32 specifically includes: Construct an equipment-parameter-cost-risk assessment model, input equipment parameters into the model, and use Monte Carlo simulation to predict the probability of the equipment meeting the standards at the effective time point.

[0046] The system monitors whether the equipment parameters meet the standards through real-time diagnostics. If they do, a change execution acceptance report is generated; if they do not, the reasons are analyzed. All change-related documents are automatically archived in the enterprise document management system, forming a traceable compliance management file.

[0047] Compliance diagnostic report generation: Report generation is tiered: corresponding reports are generated based on the compliance risk level. Emergency level (such as endangering equipment safety) generates a "red warning report", which includes real-time curves of parameters exceeding the standard, risk spread prediction (such as the value that may be reached within 10 minutes), and emergency response procedures; high level (important parameters exceeding the standard) generates an "orange warning report", which supplements the comparison of historical records of exceeding the standard; low to medium level generates a "yellow / blue warning report", which focuses on trend analysis and preventive recommendations.

[0048] Intelligent rectification suggestions: Based on the "problem-solution" association knowledge base (storing 5000+ historical rectification cases), the cosine similarity algorithm is used to match similar problems and output rectification solutions; for complex problems (such as those requiring equipment modification), the rectification costs (materials, labor, downtime losses) and compliance benefits (avoiding penalties, extending equipment life) are automatically calculated, and a cost-benefit analysis table is provided.

[0049] Multi-terminal adaptation and interaction: The report supports web (responsive layout, adapted to PC / tablet), mobile (WeChat mini program / APP) and industrial large screen display; the mobile terminal supports voice broadcast warning (for on-site operation scenarios of maintenance personnel), the web terminal provides report annotation function, and medical staff can add modification opinions and send them back to the system; the system automatically tracks the progress of the rectification plan and updates the status every 2 hours (not started / in progress / completed / delayed), and triggers a secondary warning when there is a delay.

[0050] Open data interfaces: Diagnostic report data is made available to enterprise ERP and MES systems through a RESTful API interface, supporting data linkage analysis; at the same time, standardized regulatory interface data (compliant with the National Energy Administration's "Compliance Regulatory Data Specifications for Power Generation Enterprises") is generated, which is convenient for regulatory authorities to access and view remotely.

[0051] Standard update monitoring and analysis: Connect to the official databases of the National Standardization Management Committee and industry associations (such as the China Electricity Council Standards Platform), set up standard update monitoring triggers (automatically crawl update information every morning at midnight); when a standard update is detected (such as the addition of GB 13223-2011 "Emission Standard of Air Pollutants for Thermal Power Plants"), automatically download the updated standard document, identify the changes through NLP semantic comparison algorithms (such as "particulate matter emission concentration adjusted from 30mg / m³ to 20mg / m³"), and mark the added, revised and abolished clauses.

[0052] Standard update monitoring and analysis: Multi-source monitoring channels: In addition to connecting to national / industry official databases, the following new channels have been added: a) Standard publisher subscription interface (such as the China Standards Press digital platform) to obtain standard publication announcements and official texts; b) Industry expert database push (inviting 20+ field experts to share standard revision dynamics in real time); c) Linkage with the enterprise's internal standard management system to synchronize updates of enterprise-defined technical specifications.

[0053] Update priority determination: Establish a standard update priority model, comprehensively consider the standard level (national standard > industry standard > enterprise standard), scope of impact (the number of devices involved > 100 is considered large scope), risk level (safety > environmental protection > performance), output priority (P1-P4), P1 level updates (such as major revisions to safety standards) trigger the processing process immediately, P4 level (non-critical internal enterprise specifications) can be delayed until weekly centralized processing.

[0054] Semantic parsing enhancement: Employing a "BERT+CRF" named entity recognition model (fine-tuned on an industry standard corpus, with an F1 score ≥ 0.95), key information such as "threshold type," "adjustment range," and "effective time" in the changed clauses is accurately extracted; the differences between the old and new clauses are compared using text similarity algorithms (such as SimBERT) to generate a "change comparison matrix," labeling change types such as "new parameters," "threshold adjustments," and "changes in scope of application"; for ambiguous expressions (such as "appropriately reduced"), quantitative conversion is performed by combining industry conventions and expert rule bases (such as defining "appropriately reduced" as 80%-90% of the original threshold).

[0055] Structured storage of changed content: The parsed change information is stored in JSON-LD format, including fields such as "Standard ID", "Change Clause ID", "Change Type", "Old Value", "New Value", and "Effective Conditions", which can be directly called when the knowledge graph is incrementally updated.

[0056] Incremental reconstruction of knowledge graph: Incremental update transaction management: A three-stage transaction process of "pre-update-verify-commit" is adopted. First, the update operation (adding / revising / abolishing nodes and relationships) is performed in the temporary graph branch to avoid affecting the operation of the main graph. After the update is completed, multi-dimensional verification (consistency, integrity, and rationality) is started. When the verification pass rate is ≥99%, it is committed to the main graph; otherwise, it is rolled back and an error log is generated. A version control mechanism is introduced. Each incremental update generates a new graph version (such as V2.1), and version rollback is supported (rollback response time ≤30s).

[0057] Detailed conflict resolution: In response to conflicts arising from standard changes, in addition to the existing priority principles, the following new rules are added: a) Exceptional rules for special scenarios, such as "If the equipment is under major repair, the implementation of the new standard can be temporarily suspended, and the standard must be met within one month after the repair is completed"; b) Matching of enterprise exemption clauses, if an enterprise has obtained a standard exemption document from the regulatory authority, the new standard relationship of the corresponding equipment will be automatically marked as "exempt", and the exemption period and supporting document link will be attached; c) Parameter conflict negotiation mechanism, when different standards have conflicting threshold requirements for the same parameter and there is no clear priority, the conflict information will be automatically pushed to the industry expert database, and the final threshold will be determined by expert voting (voting period ≤ 24 hours).

[0058] Distributed Restructuring Optimization: Using the Spark distributed computing framework, the incremental update task of the knowledge graph is decomposed into three parallel tasks: "node update", "relationship update" and "index reconstruction", which are distributed to different computing nodes (the number of nodes can be dynamically expanded, supporting a maximum of 10 nodes in parallel); for large-scale graphs (number of nodes > 100,000), a partition update strategy is adopted (partitioned by device type / standard category), with a single partition update time ≤ 10 minutes and the entire graph update time controlled within 1 hour.

[0059] Impact assessment and change notification: Multi-dimensional Impact Assessment Model: Construct an assessment model of "equipment-parameter-cost-risk". Input equipment parameters (current value, historical fluctuation range), standard change requirements (new threshold, effective time), and equipment attributes (remaining lifespan, modification difficulty). Output: a) Compliance Feasibility: Use Monte Carlo simulation (simulation times ≥ 1000 times) to predict the probability of equipment compliance at the effective time. Probability > 80% indicates "no modification required", 50%-80% indicates "parameter optimization required", and < 50% indicates "equipment upgrade required"; b) Modification Scheme Comparison: For equipment requiring upgrades, generate 2-3 modification schemes (e.g., "replace sensor" vs. "modify equipment components"). Compare the modification cycle (e.g., 7 days vs. 30 days), investment cost (e.g., 50,000 yuan vs. 200,000 yuan), and operation and maintenance cost (annual consumption 5,000 yuan vs. 10,000 yuan) of each scheme, and recommend the optimal scheme; c) Risk Transmission Analysis: Assess the impact of standard changes on related systems. For example, "reduction of nitrogen oxide emission limits" may lead to "increased load on the denitrification system", which in turn affects "boiler thermal efficiency". Output a risk transmission path diagram.

[0060] Tiered push strategy: Different push methods are adopted according to the urgency and scope of impact of the assessment results: P1 level updates (such as security risks) are pushed to the operation and maintenance manager and senior management of the enterprise through SMS + telephone + system pop-up; P2 level (environmental risks) are pushed to the operation and maintenance supervisor through SMS + system notification; P3-P4 level updates are pushed to the operation and maintenance engineer through system message; the push information includes "change summary", "list of affected equipment", "suggested handling time limit" and "emergency contact person" to ensure that responsibility is assigned to the right person.

[0061] Closed-loop management and effect verification: Establish a closed-loop process of "change push - solution execution - effect verification - archiving". The system automatically tracks the progress of solution execution and pushes an upgrade every 8 hours if it is not processed within the time limit. After the transformation is completed, the system monitors whether the equipment parameters meet the standards through real-time diagnostic layer. If they meet the standards, a "change execution acceptance report" is generated. If they do not meet the standards, the system analyzes the reasons and pushes secondary rectification suggestions. All change-related documents (assessment report, rectification plan, acceptance report) are automatically archived to the enterprise document management system to form a traceable compliance management file.

[0062] In summary, this invention breaks through the limitations of static standard queries: existing technologies only enable the retrieval and query of standard documents, requiring manual matching of equipment parameters; this invention achieves dynamic association between the two through a "standard-equipment" related knowledge graph, upgrading compliance diagnosis from "manual comparison" to "automatic system matching", improving efficiency by more than 30 times.

[0063] Real-time compliance risk identification: Existing compliance systems rely on offline data upload and analysis, which has a delay of several hours; this invention deploys edge computing nodes and a lightweight rule engine to achieve minute-level matching between runtime data and standards, which can provide early warning of potential compliance risks and avoid regulatory penalties.

[0064] Dynamically adapting to standard updates: Existing systems require manual reconstruction of the rule base after standard updates, which is time-consuming and labor-intensive; this invention achieves fully automated processing of standard changes through automatic monitoring, incremental reconstruction, and impact assessment, reducing the standard matching error rate by more than 90%.

[0065] Secondly, embodiments of this application provide an intelligent matching and diagnostic system 40 based on a standard-device dual knowledge base, such as... Figure 4 As shown, it includes: Knowledge fusion module 41 is used to build a standard-device dual knowledge base that includes a standard knowledge base and a device knowledge base, and to build a knowledge graph in combination with the Neo4j graph database; The real-time diagnostic module 42 is used to collect equipment operation data in real time through the edge computing node deployed on the equipment control cabinet, preprocess the equipment operation data, import the obtained equipment operation data into the lightweight rule engine for matching, and obtain the matching result. The dynamic update module 43 is used to incrementally update the knowledge graph based on the standard changes detected in real time, select the affected devices from the updated knowledge graph based on the matching results, and process them according to the type of the affected devices.

[0066] By establishing a knowledge graph linking standards and devices, dynamic correlation between the two is achieved, upgrading compliance diagnostics from manual comparison to automatic system matching, thus improving efficiency. Deploying edge computing nodes and a lightweight rule engine enables minute-level matching of operational data with standards, providing early warnings of potential compliance risks and avoiding regulatory penalties.

[0067] It should be noted that the intelligent matching and diagnostic system based on a standard-device dual knowledge base provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, the terms "module," "unit," and "subunit" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0068] Thirdly, embodiments of this application provide an electronic device, Figure 5 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 5 As shown, the electronic device may include a processor 51 and a memory 52 storing computer program instructions.

[0069] Specifically, the processor 51 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0070] The memory 52 may include a large-capacity memory for data or instructions. For example, and not limitingly, the memory 52 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 52 may include removable or non-removable (or fixed) media. Where appropriate, the memory 52 may be internal or external to a data processing device. In a particular embodiment, the memory 52 is non-volatile memory. In a particular embodiment, the memory 52 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0071] The memory 52 can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 51.

[0072] The processor 51 reads and executes computer program instructions stored in the memory 52 to implement any of the intelligent matching and diagnostic methods based on the standard-device dual knowledge base in the above embodiments.

[0073] In one embodiment, the electronic device may further include a communication interface 53 and a bus 50. Wherein, as... Figure 5 As shown, the processor 51, memory 52, and communication interface 53 are connected through bus 50 and complete communication with each other.

[0074] The communication interface 53 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication port 53 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0075] Bus 50 includes hardware, software, or both, that couples the components of the electronic device together. Bus 50 includes, but is not limited to, at least one of the following: data bus, address bus, control bus, expansion bus, and local bus. For example, and not as a limitation, bus 50 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 50 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0076] Fourthly, embodiments of this application provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the intelligent matching and diagnostic method based on a standard-device dual knowledge base provided in the first aspect.

[0077] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0078] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, is used to cause the terminal device to perform the steps of implementing the intelligent matching diagnostic method based on a standard-device dual knowledge base provided in the first aspect.

[0079] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. An intelligent matching and diagnostic method based on a standard-device dual knowledge base, characterized in that, include: Construct a standard-device dual knowledge base that includes a standard knowledge base and a device knowledge base, and build a knowledge graph by combining the Neo4j graph database; By collecting equipment operation data in real time through edge computing nodes deployed on the equipment control cabinet, the equipment operation data is preprocessed and then imported into a lightweight rule engine for matching to obtain matching results. The knowledge graph is incrementally updated based on the standard changes detected in real time. Affected devices are selected from the updated knowledge graph based on the matching results, and then processed according to the type of the affected devices.

2. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 1, characterized in that, The construction of a standard-device dual knowledge base, comprising a standard knowledge base and a device knowledge base, and the building of a knowledge graph using the Neo4j graph database, includes: Obtain national supervision standards or industry technical supervision standards, construct the standard knowledge base based on the national supervision standards, construct the equipment knowledge base based on the industry technical supervision standards, and associate the standard knowledge base and the equipment knowledge base to obtain the standard-equipment dual knowledge base; Based on the equipment knowledge base, a hierarchical relationship model containing equipment, components, rated parameters, and operation and maintenance status is constructed to build a four-element relationship model, and the multi-dimensional association rules in the four-element relationship model are determined. The knowledge graph is constructed based on the Neo4j graph database. The knowledge graph is then logically validated according to the multi-dimensional association rules. After the validation is completed, a visualized knowledge graph is generated.

3. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 2, characterized in that, The process of constructing the knowledge graph based on the Neo4j graph database, performing logical verification on the knowledge graph according to the multi-dimensional association rules, and generating a visualized knowledge graph after verification includes: The knowledge graph is constructed based on the Neo4j graph database, and the knowledge graph includes a basic layer, a relational layer, and a derivative layer. The multi-dimensional association rules are imported into the knowledge graph reasoning engine, and the knowledge graph reasoning engine is used to perform consistency verification, integrity verification and rationality verification in sequence. After the verification is completed, the visualized knowledge graph is generated.

4. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 1, characterized in that, The process involves real-time collection of equipment operation data via edge computing nodes deployed on the equipment control cabinet, preprocessing the equipment operation data, importing the obtained equipment operation data into a lightweight rule engine for matching, and obtaining matching results, including: The edge computing nodes are deployed in the equipment control cabinet. Based on the edge computing nodes, the equipment operation data is collected in real time. The equipment operation data is then preprocessed, including wavelet denoising, outlier removal, and data standardization, to obtain the processed equipment operation data. The lightweight rule engine is customized based on Drools rules. The device operation data is imported into the customized lightweight rule engine to perform matching priority management and obtain the matching result.

5. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 4, characterized in that, The process involves customizing the lightweight rule engine based on Drools rules, importing the device operation data into the customized lightweight rule engine to perform matching priority management, and obtaining the matching results, including: Based on the Drools rules, redundant modules in the lightweight rule engine are removed to obtain the customized lightweight rule engine; The lightweight rule engine performs multi-dimensional matching, including trend matching and device group matching. For scenarios where standard clauses are vaguely worded, a fuzzy membership function is constructed by combining historical equipment operation data to achieve quantitative matching of qualitative standards, thus obtaining the matching result.

6. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 1, characterized in that, The step of incrementally updating the knowledge graph based on real-time detected standard changes, selecting affected devices from the updated knowledge graph based on the matching results, and processing the affected devices according to their type includes: Based on the standard update monitoring trigger, when the standard is updated, the updated standard document is downloaded, the changes are identified, and new, revised and obsolete clauses are marked. A device-parameter-cost-risk assessment model is constructed, and Monte Carlo simulation is used to predict the probability of the device meeting the standard at the effective time point. Based on the probability of meeting the standard, affected devices are screened out, and the affected devices are processed according to different types.

7. The intelligent matching and diagnostic method based on a standard-device dual knowledge base according to claim 6, characterized in that, The constructed equipment-parameter-cost-risk assessment model uses Monte Carlo simulation to predict the probability of equipment meeting standards at the effective time point. Based on the probability of meeting standards, affected equipment is screened out, and the affected equipment is processed according to different types, including: Construct the equipment-parameter-cost-risk assessment model, input equipment parameters into the equipment-parameter-cost-risk assessment model, and use the Monte Carlo simulation to predict the probability of the equipment meeting the standards at the effective time point; The system monitors whether the equipment parameters meet the standards through real-time diagnostics. If they do, a change execution acceptance report is generated; if they do not, the reasons are analyzed. All change-related documents are automatically archived in the enterprise document management system, forming a traceable compliance management file.

8. An intelligent matching and diagnostic system based on a standard-device dual knowledge base, characterized in that, include: The knowledge fusion module is used to build a standard-device dual knowledge base, which includes a standard knowledge base and a device knowledge base, and to build a knowledge graph in conjunction with the Neo4j graph database. The real-time diagnostic module is used to collect equipment operation data in real time through edge computing nodes deployed on the equipment control cabinet, preprocess the equipment operation data, import the obtained equipment operation data into a lightweight rule engine for matching, and obtain matching results. The dynamic update module is used to incrementally update the knowledge graph based on the standard changes detected in real time, select the affected devices from the updated knowledge graph based on the matching results, and process the affected devices according to their types.

9. An electronic device, characterized in that, include memory, processor, and A computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the intelligent matching diagnostic method based on a standard-device dual knowledge base as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the intelligent matching and diagnostic method based on a standard-device dual knowledge base as described in any one of claims 1 to 7.

Images