Safety production risk conduction analysis method and device, storage medium and computer equipment

By constructing a dynamic Bayesian network model for power grid operation risk transmission analysis, the problems of delayed risk assessment and lack of foresight in the existing system are solved. It achieves deep integration of multi-source data and accurate quantification of risk levels, provides support for pre-event prediction and in-event intervention, and improves the safety of power grid operations.

CN121303848APending Publication Date: 2026-01-09GUANGZHOU JINGKAI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511530769.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-10-24
Publication Date
2026-01-09

AI Technical Summary

Technical Problem

The existing power grid operation safety access management system cannot analyze the dynamic coupling effect between elements, and the risk assessment is lagging and lacks foresight, making it impossible to carry out proactive prevention and control.

Method used

By acquiring multi-source data, a dynamic Bayesian network (DBN) model is constructed to conduct risk transmission analysis, infer the overall risk level and transmission path, and provide visualization and real-time early warning.

Benefits of technology

It achieves deep integration and correlation analysis of multi-source data such as people, plans, tools, and environment, accurately quantifies risk levels, and realizes the transformation from post-event analysis to pre-event prediction and in-event intervention, proactively reducing systemic risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121303848A_ABST
    Figure CN121303848A_ABST
Patent Text Reader

Abstract

According to the safety production risk conduction analysis method and device, the storage medium and the computer equipment provided by the invention, when the target operation task is received, the system obtains the multi-source data related to the task, including the personnel qualification data, the operation plan data, the tool data, the work ticket data and the environment data. And then, a target risk conduction model corresponding to the target operation task can be determined, and after the multi-source data is converted into standardized evidence, the overall risk level and risk value of the target operation task and a risk conduction path are reasoned based on the standardized evidence through the target risk conduction model. And finally, the risk levels, the risk values and the risk conduction paths can be visually displayed, and real-time early warning is performed according to the risk values and the conduction paths, so that a data island is broken, deep fusion and correlation analysis of multi-source data such as people, plans, tools, tickets and environments are realized, and systematic risks are actively reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of power grid safety supervision and operation technology, and in particular to a method, device, storage medium and computer equipment for analyzing the transmission of safety production risks. Background Technology

[0002] Currently, the power grid industry is vast in scale, with numerous and widespread on-site operations (such as transmission line maintenance, substation equipment pre-testing, and live-line work on distribution networks), posing extremely high safety risks. Although a mature safety management system has been established, many challenges remain regarding the real-time nature, accuracy, and systematic nature of risk warnings. For example, existing rule-based power grid operation safety access management systems primarily achieve this through the following steps:

[0003] 1. Data storage: Personnel qualifications, work plans, tools, and work permits (the “four elements”) are stored in different independent systems or database modules.

[0004] 2. Static Validation: During critical steps (such as work order issuance and tool requisition), the system executes pre-configured rule checks. For example, it checks whether the worker's qualification certificate is valid; whether the work plan has been approved; and whether the tools have passed the most recent test.

[0005] 3. Output: If the verification passes, the process continues; if it fails, an alarm will pop up and the process will be blocked.

[0006] This solution implements basic, single-point security access control, but it has the following technical problems:

[0007] 1. Isolation: Only the "four elements" are examined in an isolated and static manner, making it impossible to analyze the dynamic coupling effects between the elements.

[0008] 2. Lag: Risk assessment is based on historical data and cannot be integrated with or respond to real-time changes. For example, the weather may be good when the work plan is approved, but a sudden thunderstorm and strong winds may occur on-site during actual operations (environmental data), and the system cannot dynamically raise the risk level and issue an alarm.

[0009] 3. Lack of foresight: Unable to deduce and predict the transmission path of risks.

[0010] 4. Passive response: Alarms are usually issued at process checkpoints, which is a passive prevention and control measure. It cannot provide risk simulation and early warning during the work planning stage, so as to achieve proactive avoidance. Summary of the Invention

[0011] The purpose of this application is to at least address one of the aforementioned technical deficiencies, particularly the technical deficiencies of existing power grid operation safety access management systems, which cannot analyze the dynamic coupling effects between elements, have a lag in risk assessment and lack foresight, and are unable to conduct proactive prevention and control.

[0012] This application provides a method for analyzing the transmission of safety production risks, the method comprising:

[0013] When a target task is received, multi-source data related to the target task is acquired. The multi-source data includes personnel qualification data, work plan data, tool and equipment data, work order data, and environmental data.

[0014] Determine the target risk transmission model corresponding to the target task;

[0015] The multi-source data is converted into standardized evidence, and the overall risk level and risk value of the target task, as well as the risk transmission path, are inferred from the standardized evidence using the target risk transmission model.

[0016] The overall risk level and risk value, as well as the risk transmission path, are visualized and displayed, and real-time warnings are issued based on the risk value and the risk transmission path.

[0017] Optionally, determining the target risk transmission model corresponding to the target task includes:

[0018] Determine the attribute information of the target task;

[0019] The attribute information is matched with the pre-built DBN templates in the template library, and the target DBN template corresponding to the target job task is determined according to the matching result, wherein different job tasks correspond to different DBN templates;

[0020] The target DBN template is instantiated into a target risk transmission model for the target task.

[0021] Optionally, the DBN template construction process includes:

[0022] Determine the template to be built and the corresponding job task;

[0023] The task is decomposed into multiple time-series stages, and each time-series stage is abstracted into a time slice in the template to be constructed. Then, the risk nodes in each time slice are determined.

[0024] After establishing causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices, the DBN template is obtained.

[0025] Optionally, after establishing the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices, the DBN template is obtained, including:

[0026] Obtain a pre-built ontology of safety knowledge for power grid operations;

[0027] Based on the power grid operation safety knowledge ontology and expert experience, the causal relationship between input nodes and output nodes within the same time slice, as well as the conditional probability table of the output nodes, are determined.

[0028] Based on the aforementioned power grid operation safety knowledge ontology and expert experience, the causal relationships and conditional probability tables for the same nodes across different time slices are determined.

[0029] A DBN template is constructed based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices.

[0030] Optionally, converting the multi-source data into standardized evidence includes:

[0031] Obtain a pre-built ontology of safety knowledge for power grid operations;

[0032] Based on the power grid operation safety knowledge ontology, standardized evidence is obtained by cleaning, transforming and standardizing the multi-source data using ETL tools.

[0033] Optionally, the process of constructing the power grid operation safety knowledge ontology includes:

[0034] Based on the regulations and systems related to power grid safety production, a computer-understandable knowledge ontology for power grid operation safety is constructed.

[0035] The power grid operation safety knowledge ontology defines concepts, object attributes, data attributes, and rule axioms related to power grid safety production.

[0036] Optionally, the target risk transmission model includes multiple time slices, each time slice contains multiple risk nodes, and causal relationships and conditional probability tables are set between different nodes in the same time slice and between the same nodes in different time slices;

[0037] The step of inferring the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model includes:

[0038] The standardized evidence is input into the target risk transmission model. Based on the causal relationships and conditional probability tables between different nodes in the same time slice and between the same nodes in different time slices in the target risk transmission model, probability propagation and updates are performed. The probability value of each risk node being in a dangerous state at the current and future times is calculated. After inferring the risk transmission path of the target task, the overall risk level and risk value of the target task, as well as the risk transmission path, are output.

[0039] This application also provides a safety production risk transmission analysis device, including:

[0040] The data acquisition module is used to acquire multi-source data related to the target task when a target task is received. The multi-source data includes personnel qualification data, work plan data, tool data, work order data, and environmental data.

[0041] The model determination module is used to determine the target risk transmission model corresponding to the target task.

[0042] The risk reasoning module is used to convert the multi-source data into standardized evidence, and to infer the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence through the target risk transmission model.

[0043] The display and early warning module is used to visualize the overall risk level and risk value, as well as the risk transmission path, and to provide real-time early warnings based on the risk value and the risk transmission path.

[0044] This application also provides a computer-readable storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the safety production risk transmission analysis method as described in any of the above embodiments.

[0045] This application also provides a computer device, including: one or more processors, and memory;

[0046] The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the safety production risk transmission analysis method as described in any of the above embodiments.

[0047] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0048] The safety production risk transmission analysis method, device, storage medium, and computer equipment provided in this application, upon receiving a target work task, will acquire multi-source data related to the task, including personnel qualification data, work plan data, tool data, work permit data, and environmental data. Next, this application can determine the target risk transmission model corresponding to the target work task, and after converting the multi-source data into standardized evidence, it can infer the overall risk level and risk value of the target work task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model. Finally, this application can visualize these risk levels, risk values, and risk transmission paths, and provide real-time early warnings based on the risk values ​​and transmission paths, thereby breaking down data silos and achieving deep integration and correlation analysis of multi-source data such as personnel, plans, tools, work permits, and the environment.

[0049] Furthermore, the system has achieved a leap from static rule verification to dynamic probability assessment, accurately quantifying the overall risk level. By visually simulating the transmission path of risks in the work process chain, it identifies key risk nodes, realizing a shift from "post-event analysis" to "pre-event prediction and in-event intervention." Ultimately, it provides decision support for work plan optimization and safety measure deployment, proactively reducing systemic risks. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A flowchart illustrating a method for analyzing the transmission of safety production risks provided in this application embodiment;

[0052] Figure 2 A schematic diagram illustrating the process of determining the target risk transmission model provided in this application embodiment;

[0053] Figure 3 A schematic diagram illustrating the construction process of the DBN template provided in this application embodiment;

[0054] Figure 4 A schematic diagram of a safety production risk transmission analysis device provided in this application embodiment;

[0055] Figure 5 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0056] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0057] In one embodiment, such as Figure 1 As shown, Figure 1 This application provides a flowchart illustrating a method for analyzing the transmission of safety production risks, as exemplified by an embodiment of the present application. The present application provides a method for analyzing the transmission of safety production risks, which may include:

[0058] S110: When a target work task is received, acquire multi-source data related to the target work task. The multi-source data includes personnel qualification data, work plan data, tool and equipment data, work order data, and environmental data.

[0059] In this step, when the system receives a target work task, it can acquire multi-source data related to that task. This multi-source data includes, but is not limited to, personnel qualification data, work plan data, tool and equipment data, work order data, and environmental data. Once the system acquires the multi-source data, it can perform data fusion to conduct risk transmission analysis on the target work task.

[0060] The target operational tasks in this application refer to various on-site operational tasks involved in power grid operations, such as transmission line maintenance, substation equipment pre-testing, and live-line work on distribution networks. These tasks often involve complex operating environments and numerous personnel and equipment, and are therefore accompanied by high safety risks. By acquiring multi-source data closely related to these target operational tasks, the system can gain a more comprehensive understanding of the actual situation at the work site, providing an accurate and reliable data foundation for subsequent risk transmission analysis.

[0061] For example, when the system receives a target job task, since this target job task includes unique identification information such as work order number, plan ID, and task sequence number, this application can obtain multi-source data from various systems based on the unique identification information of this job task. An example is as follows:

[0062] 1. Personnel Qualification Evidence: The system retrieves a list of all personnel involved in the operation from the "Two-Ticket System" using the work order number. Then, using these personnel IDs, the system batch queries each person's qualifications, training, violation records, and other information in the "Personnel Skills System."

[0063] 2. Tool and Equipment Evidence: Using the work order number, retrieve the list of tools and equipment to be requisitioned for this work plan from the "Two-Ticket System" or "Tool and Equipment Management System". Then, based on these tool asset numbers, check their latest testing status and maintenance records in the "Asset Management System". It is worth noting that before the work begins, the warehouse RFID scanning confirmation will update the tool's "requisitioned, on-site" status evidence again.

[0064] 3. Work plan evidence: Obtain the detailed content, location, and time requirements of the work from the "Work Plan Management System" through the plan ID.

[0065] 4. Environmental evidence: By using the geographical location information in the work plan, the system automatically calls the meteorological bureau's API to obtain real-time weather data for that location.

[0066] Once the aforementioned multi-source data is acquired, this application can conduct risk transmission analysis on the target operation task based on the multi-source data. Risk transmission refers to the phenomenon in power grid operations where a risk factor in one stage (such as an unqualified person) can propagate along the work process chain and affect subsequent stages (such as causing operational errors, leading to safety accidents), creating a "domino effect." Therefore, this application integrates internal power grid data such as "personnel qualification data (skill library)," "operation plan data (plan library)," "tool and equipment status data (asset library)," and "work permit data (two-permit library)," as well as external environmental data (meteorological data) and real-time location data (GPS / BDS), for unified analysis to dynamically assess risks and simulate risk transmission paths.

[0067] S120: Determine the target risk transmission model corresponding to the target task.

[0068] In this step, after obtaining multi-source data related to the target task through S110, this application can determine the target risk transmission model corresponding to the target task, so that the multi-source data can be analyzed in a unified manner through the target risk transmission model.

[0069] The target risk transmission model in this application refers to a dynamic risk assessment framework constructed for different task characteristics, such as Dynamic Bayesian Network (DBN), Stochastic Petri Net (SPN), or System Dynamics (SD) models. SPN excels at describing concurrent, asynchronous discrete event systems and can be used to model risk interactions in parallel workflows; SD excels at handling high-order, nonlinear feedback problems and can be used to simulate the accumulation and amplification effects of risk factors; while DBN is more intuitive and powerful in handling uncertainty and temporal reasoning.

[0070] For example, when this application selects Dynamic Bayesian Network (DBN) as the target risk transmission model, the model uses DBN as its core and achieves dynamic simulation of the risk transmission path by quantifying the temporal dependencies and probability transmission mechanisms between various risk nodes. Specifically, the model construction process first decomposes the work process into multiple temporal stages, each corresponding to a time slice, which includes risk nodes such as personnel, equipment, and environment. Subsequently, a causal relationship table between nodes is established through expert experience and historical data. For example, "incompetent personnel qualifications" may lead to an increased probability of "operational errors," and "expired tool testing" may increase the risk of "equipment failure." At the same time, a conditional probability table across time slices is constructed to depict how risks evolve with the work process. For example, "initial personnel violations" may be transmitted to subsequent stages through "tool misuse," ultimately leading to a "safety accident." This modeling approach breaks through the limitations of traditional static rule verification and can capture the dynamic cumulative effect and cascading amplification characteristics of risks in the work chain.

[0071] S130: Convert multi-source data into standardized evidence, and infer the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model.

[0072] In this step, after determining the target risk transmission model corresponding to the target task through S120, this application can also convert multi-source data into standardized evidence. In this way, the overall risk level and risk value of the target task, as well as the risk transmission path, can be inferred from the standardized evidence based on the target risk transmission model.

[0073] For example, when multi-source data such as personnel qualification data, work plan data, tool and equipment data, work order data, and environmental data are input into the system, the system first preprocesses this data to remove noise and invalid information. Next, this application can utilize ETL tools to clean, transform, and standardize this multi-source data to obtain standardized evidence. This standardized evidence covers key information at each stage of the operation, such as personnel skill levels, equipment status, the rationality of the work plan, and environmental conditions. Subsequently, the system inputs this standardized evidence into a defined target risk transmission model. This model, using a Dynamic Bayesian Network (DBN) as an example, performs probability propagation and updates based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes across different time slices.

[0074] Through this process, the model can calculate the probability of each risk node being in a dangerous state at present and in the future, and deduce the risk transmission path of the target task. Finally, based on these calculation results, the system outputs the overall risk level and risk value of the target task, as well as a detailed risk transmission path, providing operators with comprehensive risk assessment and early warning information.

[0075] S140: Visualizes the overall risk level and risk value, as well as the risk transmission path, and provides real-time warnings based on the risk value and risk transmission path.

[0076] In this step, after inferring the overall risk level and risk value of the target task and the risk transmission path through S130, this application can visualize the overall risk level and risk value and the risk transmission path, and provide real-time early warning based on the risk value and risk transmission path.

[0077] Specifically, in this application, the system can intuitively present the overall risk level using color-coded levels (such as red, yellow, and blue) through visual charts and dynamic simulations, while the risk value is displayed in numerical and percentage form, facilitating operators' quick understanding of the current safety status of their work. Simultaneously, the risk transmission path is dynamically displayed in the form of flowcharts or network diagrams, clearly marking the propagation process of the risk from the starting node to potentially affected nodes, as well as the causal relationships and probability values ​​between each node, and highlighting the critical path.

[0078] Based on this visualized information, the system can monitor changes in risk values ​​in real time. Once the risk value exceeds the preset threshold or the risk transmission path points to a high-risk area, the early warning mechanism is immediately triggered. The system sends tiered early warning information to the person in charge and safety supervisor through various means such as sound, light, pop-up windows, text messages, and emails, indicating the risk points and suggesting corresponding safety measures, thereby effectively preventing the occurrence of safety accidents.

[0079] In the above embodiments, when a target work task is received, the system acquires multi-source data related to the task, including personnel qualification data, work plan data, tool data, work order data, and environmental data. Next, this application can determine the target risk transmission model corresponding to the target work task, and after converting the multi-source data into standardized evidence, it infers the overall risk level and risk value of the target work task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model. Finally, this application can visualize these risk levels, risk values, and risk transmission paths, and provide real-time warnings based on the risk values ​​and transmission paths, thereby breaking down data silos and achieving deep integration and correlation analysis of multi-source data such as personnel, plans, tools, work orders, and the environment.

[0080] Furthermore, the system has achieved a leap from static rule verification to dynamic probability assessment, accurately quantifying the overall risk level. By visually simulating the transmission path of risks in the work process chain, it identifies key risk nodes, realizing a shift from "post-event analysis" to "pre-event prediction and in-event intervention." Ultimately, it provides decision support for work plan optimization and safety measure deployment, proactively reducing systemic risks.

[0081] In one embodiment, such as Figure 2 As shown, Figure 2 This is a schematic diagram illustrating the process of determining the target risk transmission model provided in the embodiments of this application; S120, determining the target risk transmission model corresponding to the target task, may include:

[0082] S121: Determine the attribute information of the target task.

[0083] S122: Match the attribute information with the pre-built DBN templates in the template library, and determine the target DBN template corresponding to the target job task based on the matching results. Different job tasks correspond to different DBN templates.

[0084] S123: Instantiate the target DBN template into a target risk transmission model for the target task.

[0085] In this embodiment, when determining the target risk transmission model corresponding to the target task, the attribute information of the target task can first be clarified. This attribute information covers multiple dimensions such as task type, work environment, personnel and equipment involved, which forms the basis for building an accurate risk transmission model. Subsequently, the system intelligently matches the above attribute information with a pre-built DBN template library. The template library stores DBN templates optimized for different work scenarios. Each template is trained with historical data and verified by experts to ensure the ability to capture specific work risks.

[0086] Based on this, this application can automatically select a target DBN template that highly matches the target task according to the matching algorithm results. This template serves as a framework for risk transmission analysis, dynamically reflecting the temporal dependencies and probability propagation relationships of each risk node in the workflow. Finally, the system instantiates the selected target DBN template into a specific target risk transmission model, and through parameter configuration and data binding, makes it fully adaptable to the characteristics of the current task, thereby providing customized support for subsequent risk quantification assessment and path deduction.

[0087] In one embodiment, such as Figure 3 As shown, Figure 3 This is a schematic diagram illustrating the construction process of the DBN template provided in this application embodiment; the DBN template construction process in S122 may include:

[0088] S221: Determine the template to be built and the corresponding job task.

[0089] S222: Decompose the task into multiple time-series stages, and abstract each time-series stage into a time slice in the template to be built, and then determine the risk nodes within each time slice.

[0090] S223: After establishing the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices, the DBN template is obtained.

[0091] In this embodiment, when constructing the DBN template, this application first clarifies the template type to be constructed and the corresponding job task type, which is a prerequisite for ensuring the template's relevance and effectiveness. Subsequently, this application decomposes the selected job task into multiple sequentially related stages according to its execution flow, with each stage corresponding to a time slice in the DBN template. After determining the specific content of each time slice, the application further abstracts the risk nodes within that time slice that may affect job safety. These nodes may involve multiple aspects such as personnel operation, equipment status, and environmental conditions.

[0092] Next, this application delves into the causal relationships between different risk nodes within the same time slice, as well as the conditional probability propagation mechanism of the same risk nodes across different time slices. Through expert evaluation, historical data analysis, and simulation experiments, accurate causal relationship tables and conditional probability tables are established. These tables constitute the core of the DBN template, quantitatively describing how risks propagate and evolve within the operational process. Finally, all the above elements are integrated to form a complete DBN template. This template not only reflects the temporal characteristics of the operational tasks but also captures the dynamic interactions between risks, providing a solid foundation for subsequent risk assessment and path prediction.

[0093] In one specific implementation method, taking "110kV transmission line power outage maintenance" as an example, the standard procedure typically includes:

[0094] T1: Tool and equipment requisition and inspection;

[0095] T2: On-site safety measures setup (grounding wire installation, barrier installation, signage, etc.);

[0096] T3: Tower climbing operation;

[0097] T4: Equipment maintenance operations (replacing insulators, tightening bolts, etc.);

[0098] T5: Completion and Acceptance of Work.

[0099] Each step constitutes a time slice in the DBN. The division of time slices ensures that the risk analysis follows the logical sequence of the operations. The risk node variables within each time slice are designed around the "Man, Machine, Environment, Management" (4M) principle, combined with the specific risks of that step.

[0100] Taking the time slice "T2: On-site safety measures deployment" as an example, its risk node variables include:

[0101] Personnel qualification compliance: Whether the qualifications of the operator and the supervisor match the special operation of grounding wire connection.

[0102] Tool condition reliability: Whether the grounding wire body, insulating rod, and voltage detector are in good condition (based on the test records in the asset warehouse and the RFID scan status at the time of issuance).

[0103] Environmental suitability: Whether the current weather (such as wind speed and humidity) meets the safety regulations regarding outdoor grounding wire installation.

[0104] Operation Correctness Probability: This is an output node that depends on the states of the three input nodes mentioned above. In other words, whether the operation is correct depends on whether the person is qualified, whether the tool is usable, and whether the environment is suitable.

[0105] Safety measures completeness: This is the final state node of this stage, which is directly determined by the "probability of correct operation".

[0106] Next, this application can establish causal relationships and conditional probability tables (CPTs) between different nodes within the same time slice and between the same nodes in different time slices, based on pre-built ontology rules and expert experience. For example, a poor "tool state at time t" will lead to a decrease in the "probability of correct operation at time t", and further affect the increase in the "risk of equipment state at time t+1".

[0107] These causal relationships and conditional probability tables form the logical foundation of the DBN template, enabling the model to simulate the propagation and evolution of risk over time. For example, if in the time slice "T2: On-site Safety Measures Layout," insufficient "personnel qualification compliance" or poor "tool and equipment reliability" leads to a decrease in the "probability of correct operation," this risk may propagate to subsequent time slices, such as "T3: Tower Climbing Operation," through the "Safety Measures Completeness" node, increasing the operational risk in that stage. Specifically, if the grounding wire is not installed correctly, it may cause serious consequences such as electric shock or falls from heights during tower climbing.

[0108] Furthermore, the DBN model can capture the cascading effects of risks across different time slices. For example, a tool defect in "T1: Tool Requisition and Inspection" may indirectly affect tower climbing safety in "T3" through the "T2" operation, and even impact the accuracy of "T4: Equipment Maintenance Operation," ultimately leading to a failure in "T5: Work Completion and Acceptance." This cross-time-slice risk transmission analysis helps identify potential risk chains in advance, providing a basis for developing preventative control measures.

[0109] By constructing and applying the aforementioned DBN template, this application achieves dynamic simulation and quantitative assessment of risk transmission paths in complex work processes. Compared to traditional static analysis methods, the DBN model can more accurately reflect the temporal dependence and probabilistic transmission characteristics of risks, providing more scientific and comprehensive decision support for safe production. For example, in actual operations, the system can update model parameters based on real-time data, dynamically adjust risk warning thresholds, and ensure that risk control measures are optimized synchronously with the work process, thereby effectively reducing systemic risks and improving operational safety.

[0110] In one embodiment, after establishing the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices in S223, the DBN template is obtained, which may include:

[0111] S2231: Obtain a pre-built ontology of power grid operation safety knowledge.

[0112] S2232: Based on the power grid operation safety knowledge ontology and expert experience, determine the causal relationship between input nodes and output nodes within the same time slice, as well as the conditional probability table of the output nodes.

[0113] S2233: Based on the power grid operation safety knowledge ontology and expert experience, determine the causal relationships and conditional probability tables of the same nodes across different time slices.

[0114] S2234: Construct a DBN template based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices.

[0115] In this embodiment, acquiring a pre-built knowledge ontology for power grid operation safety is the first step in constructing the DBN template. This ontology brings together core concepts, relationships, and rules in the field of power grid operations, providing a solid knowledge foundation for risk analysis. Based on this ontology and expert experience, the system can accurately identify the causal relationships between input nodes (such as personnel qualifications and tool status) and output nodes (such as the probability of correct operation and the completeness of safety measures) within the same time slice. For example, insufficient personnel qualifications may directly lead to a decrease in the probability of correct operation; this relationship is quantified using a conditional probability table, ensuring the accuracy of risk assessment.

[0116] Simultaneously, the system also conducts in-depth analysis of the causal relationships and conditional probability tables of the same nodes across different time slices. Taking "tool status" as an example, its status in time slice T1 may affect the completeness of safety measures in time slice T2 through operational procedures, thereby impacting operational risks in subsequent time slices. This cross-time slice risk transmission analysis relies on in-depth mining of the power grid operation safety ontology and integration of expert experience, ensuring that the DBN template can comprehensively capture the propagation and evolution of risks across the time dimension.

[0117] Continuing with the example of the time slice "T2: On-site safety measures deployment," once the risk nodes within this time slice are determined, this application can define the node "Operation Correctness Probability," which is influenced by three parent nodes: Personnel Qualification Compliance (Human), Tool Status Reliability (Tool), and Environmental Suitability (Env). Furthermore, this application can define a conditional probability table for Operation. This conditional probability table is a multi-dimensional table that defines the probability that Operation is "correct" or "incorrect" under all possible combinations of parent nodes. A specific example is shown below (for simplicity, each parent node only takes two states: "yes" or "no"):

[0118] Table 1. Causal relationships and conditional probabilities among different nodes within the same time slice.

[0119]

[0120] Furthermore, since the completeness of safety measures (Safety_T2) of the final output node in stage T2 will directly serve as an input parent node for stage T3 (tower climbing operation), it can be referred to as the preliminary safety measure status (Prev_Safety_T3). The relationship between the final output node of stage T2 and the input parent node of stage T3 is typically deterministic or highly probable. This is because if safety measures are complete (grounding wire properly installed), the environmental risk of the tower climbing operation is greatly reduced; conversely, it is extremely dangerous. Therefore, this application can construct the following conditional probability table based on this causal relationship:

[0121] Table 2. Causal relationships and conditional probabilities of the same nodes across different time slices.

[0122]

[0123] In this way, the risk is shifted from "whether safety measures are in place" to "whether the environmental risk of tower climbing operations is high or low". The probability values ​​in the conditional probability table of this application were initially set by expert experience, and can be optimized later through machine learning using historical operational data (such as violation records and accident reports), without any restrictions here.

[0124] Ultimately, based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes across different time slices, the system constructs a complete DBN template. This template not only reflects the temporal characteristics of power grid operations but also quantitatively describes the dynamic interactions between risks, providing customized support for subsequent risk quantification assessment and path prediction. Through the application of the DBN template, the system can achieve dynamic simulation and quantitative assessment of risk transmission paths in complex operational processes, providing a more scientific and comprehensive basis for safe production decisions.

[0125] In one embodiment, converting the multi-source data into standardized evidence in S130 may include:

[0126] S131: Obtain a pre-built ontology of safety knowledge for power grid operations.

[0127] S132: Based on the power grid operation safety knowledge ontology, standardized evidence is obtained by cleaning, transforming and standardizing the multi-source data using ETL tools.

[0128] In this embodiment, after obtaining a pre-built knowledge ontology of power grid operation safety, the system can rely on the core concepts, relationships and rules of power grid operation covered by this ontology to perform targeted processing on multi-source data from different channels and in different formats.

[0129] Specifically, this application utilizes ETL (Extract-Transform-Load) tools to first extract the necessary key information from multi-source data. This information may include personnel operation records, equipment operating status data, environmental monitoring data, etc. Next, based on the specifications and standards defined in the power grid operation safety knowledge ontology, the extracted data is cleaned to remove noisy, erroneous, and duplicate data, ensuring data accuracy and consistency. Then, the cleaned data is transformed to conform to the format and structure required by the subsequent risk analysis model; for example, different types of data are uniformly converted into numerical data or specific encoding formats.

[0130] Finally, the data processed in this series of steps forms standardized evidence. This standardized evidence provides a reliable and unified data foundation for subsequent risk quantification assessment and path prediction based on DBN templates, making the risk analysis process more scientific and accurate, and better serving safety production decisions.

[0131] In one embodiment, the process of constructing the power grid operation safety knowledge ontology in S2231 or S131 may include:

[0132] Based on the regulations and systems related to power grid safety production, a computer-understandable power grid operation safety knowledge ontology is constructed; wherein, the power grid operation safety knowledge ontology defines concepts, object attributes, data attributes and rule axioms related to power grid safety production.

[0133] In this embodiment, constructing a computer-understandable knowledge ontology of power grid operation safety is a crucial foundation for ensuring the scientific validity and accuracy of the entire risk transmission analysis device. This process closely relies on regulations and systems related to power grid safety production, such as the power grid's "Safety Production Regulations" and "Two-Ticket Management Measures."

[0134] This application delves into and extracts the core concepts inherent in these regulations, such as "work supervisor," "special operation qualification," "Level 1 hot work permit," "insulating boots," and "expiration of testing period." Simultaneously, it clearly defines the object attributes between these concepts, such as the compatibility between equipment and work, the mitigating effect of safety measures on operational risks, "having qualifications," "requires Tool," "has Status," and "is Part of Plan." Furthermore, the knowledge ontology of this application specifies data attributes in detail, including the specific numerical range of equipment parameters, the precise interval of work time, and rule axioms, such as the combination of safety measures that must be taken under specific working conditions and the threshold at which equipment status must be reached to stop work.

[0135] Through this construction approach, the power grid operation safety knowledge ontology not only provides a comprehensive and accurate knowledge framework for risk analysis, but also ensures that subsequent risk quantification assessments and path predictions based on this ontology can closely align with the actual needs of power grid safety production, providing solid knowledge support for safety production.

[0136] In one embodiment, the target risk transmission model may include multiple time slices, each time slice having multiple risk nodes, and causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices.

[0137] S130, which uses the target risk transmission model to infer the overall risk level and risk value of the target task based on the standardized evidence, as well as the risk transmission path, may include:

[0138] The standardized evidence is input into the target risk transmission model. Based on the causal relationships and conditional probability tables between different nodes in the same time slice and between the same nodes in different time slices in the target risk transmission model, probability propagation and updates are performed. The probability value of each risk node being in a dangerous state at the current and future times is calculated. After inferring the risk transmission path of the target task, the overall risk level and risk value of the target task, as well as the risk transmission path, are output.

[0139] In this embodiment, after standardized evidence is input into the target risk transmission model, the model operates based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes across different time slices. For different nodes within the same time slice, the model updates the probability values ​​of output nodes (such as the probability of correct operation and the completeness of safety measures) in a dangerous state based on the known causal relationships and conditional probabilities, using the conditional probability tables and the status information of input nodes (such as personnel qualification compliance, equipment reliability, etc.). For example, if the personnel qualification compliance is "no", according to the conditional probability table, the probability of the correct operation being in a dangerous state (i.e., an operational error) will increase accordingly.

[0140] For the same node across different time slices, the model considers the impact of the node's state in the previous time slice on the state in the subsequent time slice. Taking the "tool / equipment state" node as an example, if the tool / equipment state is poor in time slice T1, then in the subsequent time slice T2, because the tool / equipment state may affect the operation, the probability of risk nodes related to the operation (such as the probability of correct operation) being in a dangerous state in time slice T2 will increase. Through this cross-time slice probability propagation and update mechanism, the model progressively calculates the probability value of each risk node being in a dangerous state at the current and future times.

[0141] After calculating the probability values ​​of all risk nodes, the model infers the risk transmission path of the target task based on these probability values ​​and the causal relationships between nodes. For example, an abnormality in the condition of tools and equipment in the "T1: Tool Retrieval and Inspection" stage, transmitted through the operational steps in the "T2: On-site Safety Measures Arrangement" stage, affects the operational risk in the "T3: Tower Climbing Operation" stage, forming a clear risk transmission path. Finally, the model integrates the probability values ​​of all risk nodes and the risk transmission path to output the overall risk level and risk value of the target task, providing an intuitive and crucial reference for safety production decisions.

[0142] The safety production risk transmission analysis device provided in the embodiments of this application is described below. The safety production risk transmission analysis device described below and the safety production risk transmission analysis method described above can be referred to in correspondence.

[0143] In one embodiment, such as Figure 4 As shown, Figure 4 This application provides a structural schematic diagram of a safety production risk transmission analysis device according to an embodiment of the present application. The present application also provides a safety production risk transmission analysis device, which may include a data acquisition module 210, a model determination module 220, a risk reasoning module 230, and a display and early warning module 240, specifically including the following:

[0144] The data acquisition module 210 is used to acquire multi-source data related to the target task when a target task is received. The multi-source data includes personnel qualification data, work plan data, tool data, work order data, and environmental data.

[0145] The model determination module 220 is used to determine the target risk transmission model corresponding to the target operation task.

[0146] The risk reasoning module 230 is used to convert the multi-source data into standardized evidence, and to infer the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence through the target risk transmission model.

[0147] The display and early warning module 240 is used to visualize the overall risk level and risk value, as well as the risk transmission path, and to provide real-time early warnings based on the risk value and the risk transmission path.

[0148] In the above embodiments, when a target work task is received, the system acquires multi-source data related to the task, including personnel qualification data, work plan data, tool data, work order data, and environmental data. Next, this application can determine the target risk transmission model corresponding to the target work task, and after converting the multi-source data into standardized evidence, it infers the overall risk level and risk value of the target work task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model. Finally, this application can visualize these risk levels, risk values, and risk transmission paths, and provide real-time warnings based on the risk values ​​and transmission paths, thereby breaking down data silos and achieving deep integration and correlation analysis of multi-source data such as personnel, plans, tools, work orders, and the environment.

[0149] Furthermore, the system has achieved a leap from static rule verification to dynamic probability assessment, accurately quantifying the overall risk level. By visually simulating the transmission path of risks in the work process chain, it identifies key risk nodes, realizing a shift from "post-event analysis" to "pre-event prediction and in-event intervention." Ultimately, it provides decision support for work plan optimization and safety measure deployment, proactively reducing systemic risks.

[0150] In one embodiment, this application also provides a computer-readable storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the safety production risk transmission analysis method as described in any of the above embodiments.

[0151] In one embodiment, this application also provides a computer device, including: one or more processors, and memory.

[0152] The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the safety production risk transmission analysis method as described in any of the above embodiments.

[0153] Indicatively, such as Figure 5 As shown, Figure 5 This is a schematic diagram of the internal structure of a computer device 300 provided in an embodiment of this application. The computer device 300 can be provided as a server. (Refer to...) Figure 5 The computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by memory 301 for storing instructions, such as application programs, that can be executed by the processing component 302. The application programs stored in memory 301 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 302 is configured to execute instructions to perform the safety production risk transmission analysis method of any of the above embodiments.

[0154] The computer device 300 may also include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate on an operating system stored in memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0155] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0156] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0157] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0158] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for analyzing the transmission of safety production risks, characterized in that, The method includes: When a target task is received, multi-source data related to the target task is acquired. The multi-source data includes personnel qualification data, work plan data, tool and equipment data, work order data, and environmental data. Determine the target risk transmission model corresponding to the target task; The multi-source data is converted into standardized evidence, and the overall risk level and risk value of the target task, as well as the risk transmission path, are inferred from the standardized evidence using the target risk transmission model. The overall risk level and risk value, as well as the risk transmission path, are visualized and displayed, and real-time warnings are issued based on the risk value and the risk transmission path.

2. The method for analyzing the transmission of safety production risks according to claim 1, characterized in that, The determination of the target risk transmission model corresponding to the target task includes: Determine the attribute information of the target task; The attribute information is matched with the pre-built DBN templates in the template library, and the target DBN template corresponding to the target job task is determined according to the matching result, wherein different job tasks correspond to different DBN templates; The target DBN template is instantiated into a target risk transmission model for the target task.

3. The method for analyzing the transmission of safety production risks according to claim 2, characterized in that, The construction process of the DBN template includes: Determine the template to be built and the corresponding job task; The task is decomposed into multiple time-series stages, and each time-series stage is abstracted into a time slice in the template to be constructed. Then, the risk nodes in each time slice are determined. After establishing causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices, the DBN template is obtained.

4. The method for analyzing the transmission of safety production risks according to claim 3, characterized in that, After establishing the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices, a DBN template is obtained, including: Obtain a pre-built ontology of safety knowledge for power grid operations; Based on the power grid operation safety knowledge ontology and expert experience, the causal relationship between input nodes and output nodes within the same time slice, as well as the conditional probability table of the output nodes, are determined. Based on the aforementioned power grid operation safety knowledge ontology and expert experience, the causal relationships and conditional probability tables for the same nodes across different time slices are determined. A DBN template is constructed based on the causal relationships and conditional probability tables between different nodes within the same time slice and between the same nodes in different time slices.

5. The method for analyzing the transmission of safety production risks according to claim 1, characterized in that, The step of converting the multi-source data into standardized evidence includes: Obtain a pre-built ontology of safety knowledge for power grid operations; Based on the power grid operation safety knowledge ontology, standardized evidence is obtained by cleaning, transforming and standardizing the multi-source data using ETL tools.

6. The method for analyzing the transmission of safety production risks according to claim 4 or 5, characterized in that, The construction process of the power grid operation safety knowledge ontology includes: Based on the regulations and systems related to power grid safety production, a computer-understandable knowledge ontology for power grid operation safety is constructed. The power grid operation safety knowledge ontology defines concepts, object attributes, data attributes, and rule axioms related to power grid safety production.

7. The method for analyzing the transmission of safety production risks according to claim 1, characterized in that, The target risk transmission model includes multiple time slices, each time slice contains multiple risk nodes, and causal relationships and conditional probability tables are set between different nodes in the same time slice and between the same nodes in different time slices. The step of inferring the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence using the target risk transmission model includes: The standardized evidence is input into the target risk transmission model. Based on the causal relationships and conditional probability tables between different nodes in the same time slice and between the same nodes in different time slices in the target risk transmission model, probability propagation and updates are performed. The probability value of each risk node being in a dangerous state at the current and future times is calculated. After inferring the risk transmission path of the target task, the overall risk level and risk value of the target task, as well as the risk transmission path, are output.

8. A safety production risk transmission analysis device, characterized in that, include: The data acquisition module is used to acquire multi-source data related to the target task when a target task is received. The multi-source data includes personnel qualification data, work plan data, tool data, work order data, and environmental data. The model determination module is used to determine the target risk transmission model corresponding to the target task. The risk reasoning module is used to convert the multi-source data into standardized evidence, and to infer the overall risk level and risk value of the target task, as well as the risk transmission path, based on the standardized evidence through the target risk transmission model. The display and early warning module is used to visualize the overall risk level and risk value, as well as the risk transmission path, and to provide real-time early warnings based on the risk value and the risk transmission path.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the safety production risk transmission analysis method as described in any one of claims 1 to 7.

10. A computer device, characterized in that, include: One or more processors, and memory; The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the safety production risk transmission analysis method as described in any one of claims 1 to 7.