Power aerial work falling risk dynamic assessment and early warning method and system

Abstract

The invention relates to the technical field of electric power system early warning, and particularly discloses an electric power high-altitude operation falling risk dynamic assessment and early warning method and system, and the method comprises the steps: obtaining real-time personnel state data, equipment state data, environment data and operation flow data of a plurality of operation links in an electric power high-altitude operation process; extracting a risk factor corresponding to each operation link according to the real-time personnel state data and the equipment state data, and obtaining an initial risk feature vector of each operation link according to the risk factors; and dynamically correcting the initial risk feature vector according to the environment data and the operation process data to obtain a calibration risk feature vector corresponding to each operation link. Through multi-source real-time data acquisition (personnel state, equipment state, environment and process data), the problems of single dimension and poor timeliness of traditional assessment data are solved, and a comprehensive and dynamic data basis is provided for risk assessment.

Description

Technical Field

[0001] This invention relates to the field of power system early warning technology, and in particular to a method and system for dynamic assessment and early warning of fall risk during high-altitude power operations. Background Technology

[0002] In the field of power system operation and maintenance and construction, high-altitude power operations (such as pole climbing and maintenance of transmission line towers, high-altitude maintenance of substation equipment, and pole climbing operations of distribution network lines) are the core types of operations to ensure the stable operation of the power grid. These operations often involve heights of more than 10 meters and need to be carried out in confined spaces such as poles, crossarms, and work platforms. The operation process is easily affected by multiple factors such as the natural environment, equipment status, and personnel operation.

[0003] Existing technologies often focus on personnel qualifications (such as whether they hold a high-altitude work permit) or static equipment records (such as the service life of towers), ignoring dynamic environmental data (such as sudden gusts of wind and real-time temperature) and work process data (such as supervisors leaving their posts and exceeding work time limits). Moreover, they mostly adopt a "one-time data collection before work" model, which cannot capture real-time changes in risk factors during work (such as a sudden increase in the worker's heart rate or loosening of tower bolts), resulting in risk perception lagging behind actual working conditions. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for dynamic assessment and early warning of fall risk during high-altitude power operations, in order to solve the technical problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for dynamic assessment and early warning of fall risk during high-altitude power operations, comprising: Acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple operational stages during high-altitude power operations; Based on the real-time personnel status data and equipment status data, extract the risk factors corresponding to each work step, and obtain the initial risk feature vector of each work step based on the risk factors. Based on the environmental data and work process data, the initial risk feature vector is dynamically corrected to obtain the calibration risk feature vector corresponding to each work step. Based on the real-time data, the risk transmission path between each work stage and the cross-stage correlation strength between adjacent work stages are obtained. A high-altitude operation fall risk correlation diagram is generated by combining the risk transmission path and the cross-stage correlation strength. The overall fall risk value of the operation is obtained based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. Determine whether the overall fall risk value exceeds a preset risk threshold; If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

[0006] Preferably, the step of extracting risk factors corresponding to each work stage based on the real-time personnel status data and equipment status data, and obtaining the initial risk feature vector of each work stage based on the risk factors, includes: Based on the real-time personnel status data, obtain the basic attribute data of the personnel; Based on the aforementioned basic personnel attribute data, obtain personnel risk correlation data; Obtain a set of personnel risk factors based on the aforementioned personnel risk correlation data; Obtain basic device attribute data based on the real-time device status data; Obtain equipment risk association data based on the aforementioned basic equipment attribute data; Obtain the set of equipment risk factors based on the equipment risk correlation data; The initial risk feature vectors for each operational stage are obtained based on the personnel risk factor set and the equipment risk factor set.

[0007] Preferably, the step of dynamically correcting the initial risk feature vector based on the environmental data and work process data to obtain the calibrated risk feature vector corresponding to each work step includes: Obtain basic environmental disturbance data based on environmental data; Environmental risk correlation data are obtained based on basic environmental disturbance data. Obtain environmental remediation correlation data based on environmental risk correlation data; Obtain basic operational data of the process based on the workflow data; Obtain process risk-related data based on the basic process operation data; Obtain process correction-related data based on process risk-related data; Based on the environmental correction correlation data and process correction correlation data, the initial risk feature vector is adjusted for dimension adaptation to obtain the calibration risk feature vector corresponding to each operation link.

[0008] Preferably, the step of obtaining the risk transmission path between each work stage and the cross-stage correlation strength between adjacent work stages based on the real-time data, and generating a high-altitude operation fall risk correlation map by combining the risk transmission path and the cross-stage correlation strength, includes: Based on the real-time data, obtain the basic time sequence data of the work process; Based on the time sequence data of the described work process, obtain the link-related data of each stage; Initial risk transmission path data is obtained based on the data linking the various stages; Based on real-time data acquisition, risk interaction data is obtained in each stage; Obtain raw data on the strength of risk correlation in each stage based on the basic data on risk interaction in each stage; Obtain cross-stage correlation strength adaptation data based on the original data of stage risk correlation strength; A fall risk correlation diagram for high-altitude operations is generated based on the initial risk transmission path data and the cross-stage correlation strength adaptation data.

[0009] Preferably, the step of obtaining the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram includes: Based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram, obtain the risk dimension data of each link and the core data of the graph structure. Based on the aforementioned risk dimension data and graph structure core data, obtain local risk characterization data for each stage; Risk transmission impact data are obtained based on local risk characterization data and core graph structure data. Comprehensive risk data for each stage is obtained based on risk transmission impact data and local risk characterization data for each stage. Based on the node relationships in the comprehensive risk data of each process and the core data of the graph structure, obtain the risk linkage data of the entire process; Obtain overall risk aggregation base data based on full-process risk linkage data; Based on the overall risk aggregation data, key risk information is structured and integrated to obtain the overall fall risk value for the operation.

[0010] Preferably, the step of generating graded early warning information including risk level and control strategy based on the overall fall risk value includes: Based on the overall fall risk value data, obtain basic risk characterization data; Data for determining risk levels is obtained based on basic risk characterization data; Risk warning level data is obtained based on the risk level determination criteria. Based on the risk warning level data and the previously obtained high-altitude operation fall risk correlation map, obtain the risk source and transmission path correlation data; Based on the correlation data between risk sources and transmission paths, obtain basic data for appropriate control strategies; Based on the basic data of the adaptive control strategy and the risk warning level data, obtain the structured component data of the graded early warning information; Based on the structured component data of the graded early warning information, the risk level and control strategy corresponding to each component data are obtained from the graded early warning information.

[0011] This invention also provides a dynamic assessment and early warning system for fall risks during high-altitude power operations, comprising: The real-time data acquisition module is used to acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple work stages during high-altitude power operations. The initial risk feature vector acquisition module is used to extract the risk factors corresponding to each operation step based on the real-time personnel status data and equipment status data, and to obtain the initial risk feature vector of each operation step based on the risk factors. The calibration risk feature vector acquisition module is used to dynamically correct the initial risk feature vector based on the environmental data and work process data to obtain the calibration risk feature vector corresponding to each work step. The fall risk association diagram generation module is used to obtain the risk transmission path between each operation stage and the cross-stage association strength between adjacent operation stages based on the real-time data, and generate a high-altitude operation fall risk association diagram by combining the risk transmission path and the cross-stage association strength. The fall risk value acquisition module is used to acquire the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. The judgment module is used to determine whether the overall fall risk value exceeds a preset risk threshold; If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

[0012] Preferably, the initial risk feature vector acquisition module includes: A personnel basic attribute data acquisition unit is used to acquire personnel basic attribute data based on the real-time personnel status data. The associated data acquisition unit is used to acquire personnel risk associated data based on the personnel basic attribute data; A personnel risk factor set acquisition unit is used to acquire a personnel risk factor set based on the personnel risk association data. The device basic attribute data acquisition unit is used to acquire device basic attribute data based on the real-time device status data. The equipment risk association data acquisition unit is used to acquire equipment risk association data based on the equipment basic attribute data; The equipment risk factor set acquisition unit is used to acquire the equipment risk factor set based on the equipment risk association data. The initial risk feature vector acquisition unit is used to acquire the initial risk feature vector of each operation link based on the personnel risk factor set and the equipment risk factor set.

[0013] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described dynamic assessment method for fall risk in high-altitude power operations.

[0014] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described dynamic assessment method for fall risk in high-altitude power operations.

[0015] The beneficial effects of this application are as follows: This invention solves the problems of single data dimensions and poor timeliness in traditional assessments by acquiring multi-source real-time data (personnel status, equipment status, environment, and process data), providing a comprehensive and dynamic data foundation for risk assessment. Through initial risk feature vector extraction and dynamic correction of environmental and process data, risk characterization is optimized, avoiding the disconnect between static assessments and actual working conditions, and improving the accuracy of single-stage risk characterization. A high-altitude operation fall risk correlation diagram is constructed to quantify the risk transmission path and intensity across stages, breaking through the limitations of traditional isolated assessments and achieving full-process risk linkage perception. Based on calibration vectors and correlation diagrams, the overall risk value is calculated, integrating local risks and transmission effects to ensure that the overall risk quantification aligns with actual synergistic effects. A graded early warning system containing risk levels and targeted control strategies is generated, solving the problem of traditional early warnings lacking specificity and achieving a closed loop of "assessment-early warning-control". The system modules have clear division of labor, and the method can be implemented using computer equipment and storage media, effectively reducing the risk of falls during high-altitude power operations and ensuring the safety of workers and the stable operation of the power grid. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a method flow according to an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of the system structure according to an embodiment of the present invention.

[0018] Figure 3 This is a schematic diagram of the internal structure of a computer device according to an embodiment of this application.

[0019] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0021] like Figure 1 As shown, this application provides a method for dynamic assessment and early warning of fall risk during high-altitude power operations, including: S1. Acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple operational stages during high-altitude power operations; S2. Based on the real-time personnel status data and equipment status data, extract the risk factors corresponding to each work step, and obtain the initial risk feature vector of each work step based on the risk factors. S3. Based on the environmental data and work process data, the initial risk feature vector is dynamically corrected to obtain the calibration risk feature vector corresponding to each work step. S4. Based on the real-time data, obtain the risk transmission path between each operation stage and the cross-stage correlation strength between adjacent operation stages, and generate a high-altitude operation fall risk correlation diagram by combining the risk transmission path and cross-stage correlation strength. S5. Obtain the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram; S6. Determine whether the overall fall risk value exceeds a preset risk threshold; If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

[0022] As described in steps S1-S6 above, this invention acquires a data foundation (including personnel, equipment, environment, and processes) through multi-source sensing devices. The risk of falls during high-altitude power operations stems from the dynamic influence of personnel, equipment, environment, and processes across multiple dimensions, and this risk can be transmitted between different operational stages. Only by comprehensively acquiring all four types of data in real time can biased risk perception be avoided. Data is collected through multiple sources, including IoT sensors, computer vision, positioning systems, and production management systems: real-time personnel status data comes from smart wearable devices, vision systems, and personnel management systems, covering physiological status, operational behavior, and qualification compliance; real-time equipment status data comes from IoT sensors and equipment ledgers, covering structural integrity and functional effectiveness; environmental data comes from portable weather stations and GPS, covering meteorological conditions and terrain features; and operational process data comes from production management systems and positioning systems, covering step execution and personnel configuration. Data collection covers the entire process of pole climbing, crossarm operation, conductor and ground wire operation, and pole descent, updated every 1-5 minutes to ensure complete data dimensions, timely delivery, and no omissions in any stage.

[0023] Transforming real-time personnel and equipment data into structured risk representations is crucial, as personnel capabilities and behaviors, and equipment structure and function are the core sources of fall risk. Relevant risk factors need to be extracted and integrated into vectors to make the risk calculable. Existing technologies suffer from limitations such as empirical risk factor extraction, unstructured representation, and insufficient focus on specific stages, relying on subjective judgment or textual descriptions and failing to adapt to the characteristics of different work stages. This paper adopts a hierarchical extraction, stage-adapted, and structured integration approach: First, basic personnel attribute data is filtered from real-time personnel status data, redundant information is removed, and personnel risk-related data is extracted and then categorized and quantified to form a set of personnel risk factors. Similarly, basic equipment attribute data is obtained from real-time equipment status data, equipment risk-related data is extracted and transformed into a set of equipment risk factors. Finally, combined with weights determined from historical accident data, the two sets of factors are quantified and integrated into an initial risk feature vector. During the process, the emphasis is on factors tailored to different work stages; for example, in pole climbing, the focus is on equipment factors related to foot straps, and in crossarm operations, the focus is on personnel factors related to safety belts, ensuring that the vector matches the characteristics of the stage and solving the problems of vague risk representation and stage disconnect in traditional methods.

[0024] By incorporating dynamic environmental and process disturbances, the initial risk feature vector is optimized to accurately reflect actual operational risks. Physically, changes in environmental conditions and deviations in operational processes alter the risk status of personnel and equipment in real time. The initial vector does not cover these dynamic factors and requires dynamic correction to avoid lag in risk assessment. Existing technologies suffer from limitations such as single correction dimensions, static methods, disconnected processes, and a black-box logic, for example, considering only wind speed and using fixed coefficients for correction. This step adopts a two-dimensional correction, dynamic coefficient, and process-sensitive adaptation scheme: Environmental basic disturbance data and environmental risk correlation data are sequentially obtained from environmental data to calculate dynamic environmental correction coefficients; process basic operation data and process risk correlation data are extracted from operational process data to determine dynamic process correction coefficients; combined with the sensitivity of each process to disturbances, each dimension of the initial vector is adapted and corrected, with strongly correlated dimensions directly multiplied by the correction coefficient, weakly correlated dimensions multiplied by a discount value of the coefficient, and irrelevant dimensions remaining unchanged; finally, a calibrated risk feature vector is obtained. For sudden scenarios (such as sudden gusts of wind), the correction coefficients can be adjusted in real time, and the source of correction data and calculation logic are clearly defined, ensuring the correction process is transparent and traceable, solving the problems of incompleteness, low accuracy, and lack of traceability in traditional correction methods.

[0025] By clarifying the transmission patterns of risks between operational stages, a full-process risk correlation model is constructed. The physical significance lies in the fact that risks in each stage of high-altitude power operations are not isolated; risks from preceding stages are transmitted to subsequent stages through specific paths, and the intensity of transmission varies with the correlation between stages. This relationship needs to be quantified through a correlation graph. Existing technologies neglect issues such as cross-stage risk transmission, path ambiguity, and static intensity. For example, assessing only the risk of a single stage does not clearly define the risk transmission path. This step achieves this through time-series analysis and interactive data mining: basic time-series data of the operational process is obtained from real-time data; stage-connection correlation data is filtered and transformed into initial risk transmission path data; simultaneously, basic data on stage-risk interactions is extracted from real-time data, raw data on stage-risk correlation strength is extracted, and cross-stage correlation strength adaptation data is obtained by combining dynamic correction coefficients; finally, using operational stages as nodes, risk transmission paths as directed edges, and cross-stage correlation strength as edge weights, a high-altitude operation fall risk correlation graph is constructed, and topology verification is performed to remove redundant closed-loop paths. The correlation diagram can clearly show the path and intensity of equipment defects in the pole climbing stage to the crossarm operation, providing a cross-stage correlation basis for subsequent overall risk value calculation and solving the problem of isolated risk assessment in traditional methods.

[0026] By integrating the risks at each stage and their cross-stage correlations, an overall value reflecting the risk level of the entire process is calculated. Physically, the overall fall risk of an operation is the sum of the risks at each stage and the risk transmission effects between stages. A comprehensive quantitative result of the overall risk can only be obtained by combining the stage risk representation of the calibration vector with the transmission relationships in the correlation graph. Existing technologies suffer from drawbacks such as calculating stage risks in isolation, ignoring correlation effects, and failing to aggregate values, such as simply adding the risks of each stage together. This step adopts a hierarchical aggregation and correlation fusion approach: extracting stage risk dimension data and core graph structure data from the correlation graph between the calibration risk feature vector and the high-altitude operation fall risk; matching nodes with risk dimensions to obtain local risk representation data for each stage; analyzing the risk transmission impact using the edge weights of the correlation graph to generate risk transmission impact data; fusing local risks and transmission impacts to obtain comprehensive stage risk data; connecting comprehensive risks according to the node relationships in the correlation graph to obtain full-process risk linkage data; filtering key node risk information to form the basic data for overall risk aggregation; and finally, integrating key risk information based on aggregation rules verified by historical cases to obtain the overall fall risk value for the operation. The process incorporates risk transmission attenuation coefficient and volatility coefficient to ensure that the overall value reflects risk linkage and dynamic changes, thus solving the problems of incomplete overall risk assessment and low accuracy in traditional methods.

[0027] By using threshold judgment and hierarchical control, precise early warning and handling of risks can be achieved. The physical significance lies in the need to differentiate risk levels through preset thresholds, matching appropriate control strategies to different levels to ensure that early warnings are both scientific and operable, avoiding over-warning or under-warning. Existing technologies suffer from shortcomings such as a single threshold, lack of hierarchical warnings, and lack of adaptive strategies; for example, setting only a single threshold without targeted strategies after an early warning. This step adopts a threshold-based hierarchical, risk-source-tracing, and strategy-matching approach: First, the risk warning level is determined based on the overall fall risk value and preset thresholds (set based on safety regulations and historical cases); then, combined with a high-altitude operation fall risk correlation diagram, the core risk sources and risk transmission paths leading to the current level are traced; control measures corresponding to the risk level and core risk sources are matched from a preset control strategy database; risk level, risk source, transmission path, and control measures are integrated to generate structured hierarchical early warning information; early warning information is simultaneously pushed through on-site audible and visual alarms and a mobile APP, recording the early warning push and response time; if the response times out, the warning level is automatically upgraded; after on-site control measures are implemented, the overall risk value is recalculated, and the warning stops when it falls below the threshold. Tiered early warning ensures differentiated handling for different risk levels. For example, high risk triggers personnel evacuation, while medium risk strengthens monitoring, thus solving the problems of poor targeting and low handling efficiency of traditional early warning systems.

[0028] In one embodiment, the step of extracting risk factors corresponding to each work step based on the real-time personnel status data and equipment status data, and obtaining the initial risk feature vector of each work step based on the risk factors, includes: S201. Obtain basic personnel attribute data based on the real-time personnel status data; S202. Obtain personnel risk association data based on the aforementioned personnel basic attribute data; S203. Obtain a set of personnel risk factors based on the personnel risk correlation data; S204. Obtain basic device attribute data based on the real-time device status data; S205. Obtain equipment risk association data based on the equipment basic attribute data; S206. Obtain the equipment risk factor set based on the equipment risk association data; S207. Obtain the initial risk feature vector for each operational step based on the personnel risk factor set and the equipment risk factor set.

[0029] As described in steps S201-S207 above, this invention extracts personnel risk-related data based on the personnel basic attribute data output in step S201 through risk dimension mapping and key information filtering. First, a correspondence is established between personnel basic attribute dimensions and fall risk dimensions, such as qualification attributes corresponding to qualification compliance risk, and experience attributes corresponding to experience suitability risk. Then, based on this correspondence, information in the basic attribute data that has a direct causal relationship with fall risk is filtered, eliminating irrelevant details such as specific incorrect training and assessment questions and the specific number of months of work experience. Finally, the filtering results are verified for risk correlation to ensure that all retained information directly affects fall safety during high-altitude operations. This further focuses on the core risks from structured basic attribute data, avoiding insufficient targeting due to information redundancy during subsequent risk factor extraction. Taking power transmission line pole climbing operations as an example, correlation data such as qualification matching with the job, whether experience meets the job complexity requirements, whether training is up to standard, and whether physiological state is suitable for high-altitude operations can be extracted from personnel basic attributes. The technical effect is reflected in the accurate positioning of risk-related information at the personnel level, providing a clear basis for subsequently constructing targeted risk factors.

[0030] By classifying and quantifying risk, the personnel risk correlation data output in step S202 is transformed into a set of personnel risk factors. First, based on the dimensions of the personnel risk correlation data, four independent risk factors are defined: qualification compliance, experience suitability, training adequacy, and physiological suitability, avoiding factor overlap. Then, based on power safety regulations and statistical data on falls from heights over the past five years, quantitative assignment rules are formulated, converting the qualitative descriptions of the risk correlation data into corresponding numerical values ​​(numerical ranges corresponding to risk levels, with a gradient distribution from no risk to extremely high risk). Finally, the assigned factors are validated to ensure consistency between the factor values ​​and the risk correlation data. This transformation of qualitative personnel risk correlation information into quantitative risk factors makes personnel risk calculable, providing a basic unit for subsequent risk vector construction. Taking power transmission line pole climbing operations as an example, the correlation data of "qualification matching, sufficient experience, training adequacy, and normal physiological condition" can be transformed into a set of factors corresponding to the no-risk level. The technical effect is reflected in the quantitative representation of personnel risk, ensuring that risk information can be integrated through mathematical operations to improve assessment accuracy.

[0031] Using multi-source sensing data from power aerial work platforms as input, basic equipment attribute data is extracted through dimensional segmentation and data structuring. Real-time equipment status data comes from IoT sensors (collecting bolt torque, vibration frequency, etc.), equipment ledger systems (recording type, design parameters, maintenance records), and visual inspection equipment (identifying rust, cracks, etc.). The screening process focuses on four core dimensions: equipment type, status, lifespan, and function. Redundant information such as equipment manufacturer and batch number is removed, while structured information such as specific equipment type and specifications, real-time operating parameters, years of use and maintenance status, and core functional effectiveness indicators are retained. Simultaneously, data from different sources is standardized in format (e.g., bolt torque is expressed as a percentage of the design value), and data integrity is verified, triggering on-site supplementary testing when key parameters are missing. This process transforms multi-channel, multi-format real-time equipment data into structured basic attribute information, providing data support for subsequent analysis of equipment risk-related content. Taking the poles and foot clamps used for pole climbing operations on transmission lines as an example, after screening, basic attributes such as equipment type, real-time structural parameters, service life and maintenance records, and protection function indicators can be obtained. The technical effect is reflected in removing irrelevant information from the equipment data, ensuring that the retained content is directly related to the safe operation of the equipment, and providing a reliable basis for equipment risk analysis.

[0032] Based on the basic equipment attribute data output in step S204, risk-related data is extracted through risk dimension mapping and key information filtering. First, a correspondence is established between basic equipment attribute dimensions and fall risk dimensions, such as type attributes corresponding to equipment compatibility risk, and status attributes corresponding to structural stability risk. Then, based on this correspondence, information in the basic attribute data that has a direct causal relationship with fall risk is filtered, eliminating irrelevant details such as equipment height and production materials. Finally, the filtering results are verified for risk correlation to ensure that all retained information directly affects fall safety during high-altitude operations. This approach further focuses on the core risks from structured basic equipment attribute data, avoiding insufficient targeting due to information redundancy during subsequent equipment risk factor extraction. Taking the poles and foot clamps used in power line climbing operations as an example, correlation data such as equipment type and operational scenario compatibility, whether structural parameters meet standards, whether use and maintenance are compliant, and whether core protective functions are normal can be extracted from the basic equipment attributes. The technical effect is reflected in the accurate positioning of risk-related information at the equipment dimension, providing a clear basis for subsequently constructing targeted equipment risk factors.

[0033] Through risk classification and quantitative assignment, the equipment risk-related data output in step S205 is transformed into a set of equipment risk factors. First, based on the dimensions of the equipment risk-related data, four independent risk factors are defined: equipment adaptability, structural stability, usage compliance, and functional effectiveness, avoiding factor overlap. Then, based on the power supply company's operational risk assessment specifications and statistical data on equipment-related fall accidents over the past five years, quantitative assignment rules are formulated, converting the qualitative descriptions of the risk-related data into corresponding numerical values ​​(numerical ranges corresponding to risk levels, with a gradient distribution from no risk to extremely high risk). Finally, the assigned factors are validated to ensure consistency between the factor values ​​and the risk-related data. This transformation of qualitative equipment risk-related information into quantitative risk factors makes equipment risk calculable and forms a shared quantitative system with personnel risk factors. Taking the poles and foot clamps used for transmission line climbing operations as an example, the associated data of "equipment adaptability, structural compliance, usage compliance, and normal function" can be transformed into a set of factors corresponding to the no-risk level. The technical effect is reflected in the quantitative representation of equipment risk, ensuring that subsequent calculations can be unified with personnel risk factors, thus improving the scientific nature of the overall risk assessment.

[0034] Through weight allocation and structured integration, the personnel risk factor set and the equipment risk factor set are transformed into initial risk feature vectors adapted to each operational stage. First, based on the risk contribution ratio of personnel and equipment factors in fall accidents over the past five years, the total weights of the personnel risk factor set and the equipment risk factor set are determined, and the weights of each sub-factor within these two sets are further subdivided. Simultaneously, the weights are fine-tuned for different operational stages (such as pole climbing and crossarm operations) to strengthen the impact of key factors in each stage (e.g., increasing the weight of equipment adaptability factors in pole climbing). Then, the quantified value of each risk factor is multiplied by its corresponding weight to obtain a weighted risk value. These weighted values ​​are then integrated according to a fixed dimensional order (personnel factors first, equipment factors second) to form an initial vector, which is then standardized to ensure dimensional balance. Finally, the adaptability of the vector to the operational stage is verified, and weights are readjusted if deviations occur. This integration of dispersed personnel and equipment risk factors into a unified structured risk representation provides a standardized input for subsequent dynamic correction based on environmental and process data. Taking power transmission line pole climbing operations as an example, the integrated vector can intuitively reflect the overall risk level of the "personnel-equipment" dimension of this link. The technical effect is reflected in the realization of unified quantitative representation of risk and its adaptation to the characteristics of the link, laying an accurate foundation for subsequent full-process risk assessment.

[0035] In one embodiment, the step of dynamically correcting the initial risk feature vector based on the environmental data and work process data to obtain the calibrated risk feature vector corresponding to each work step includes: S301. Obtain basic environmental disturbance data based on environmental data; S302. Obtain environmental risk correlation data based on basic environmental disturbance data. S303. Obtain environmental remediation correlation data based on environmental risk correlation data; S304. Obtain basic operational data of the process based on the work process data; S305. Obtain process risk-related data based on the basic process operation data; S306. Obtain process correction related data based on process risk related data; S307. Adjust the dimensions of the initial risk feature vector according to the environmental correction correlation data and the process correction correlation data to obtain the calibration risk feature vector corresponding to each operation link.

[0036] As described in steps S301-S307 above, this invention uses environmental data collected by multi-source environmental sensing technology at power high-altitude operation sites as input. Through dimensional segmentation, data filtering, and format standardization, it extracts basic environmental interference data. The environmental data originates from portable weather stations deployed at the operation site (collecting real-time meteorological parameters), GPS positioning and terrain databases (obtaining the type of work area and terrain risk characteristics), and surrounding environmental monitoring equipment (monitoring nearby obstacles, etc.). The filtering process focuses on two core dimensions: meteorological interference and terrain and surrounding interference. Redundant information such as terrain data from non-operation areas and historical meteorological data is eliminated. Real-time meteorological parameters, work area type, terrain risk level, and surrounding obstacle conditions related to fall risk are retained. Simultaneously, data from different sources are standardized into a uniform format (e.g., terrain type coding), and the integrity of the filtered data is verified. A supplementary data collection mechanism is triggered when core dimensions are missing. This transforms scattered, multi-format environmental data into structured basic interference information, laying the data foundation for subsequent focus on environmental risk-related content. Taking mountaintop climbing operations as an example, basic interference data such as real-time weather conditions at the work site, mountaintop terrain features, and the absence of obstacles in the surrounding area can be filtered out. The technical effect is reflected in the effective removal of irrelevant information from the environmental data, ensuring that the retained content is directly related to the risk of falling from heights, and avoiding interference from redundant data in subsequent risk analysis.

[0037] Based on the environmental basic disturbance data output in step S301, environmental risk-related data is extracted through risk dimension mapping, key information screening, and correlation verification. First, a correspondence is established between the environmental basic disturbance dimension and the fall risk dimension, clarifying that meteorological parameters correspond to personnel balance and physiological fatigue risks, terrain type corresponds to environmental amplification risks, and surrounding obstacles correspond to operational interference risks. Then, based on this correspondence, information in the basic disturbance data that has a direct causal relationship with fall risk is screened, eliminating irrelevant details such as sufficient daytime sunlight intensity. Finally, the screening results are verified for risk correlation to ensure that all retained information directly affects fall safety during high-altitude operations. This approach further focuses on the core risks from the structured environmental basic disturbance data, avoiding insufficient targeting due to information redundancy during subsequent environmental correction correlation data extraction. Taking mountaintop climbing operations as an example, correlation data such as the impact of real-time weather on personnel balance and the amplification effect of mountaintop terrain on risks can be extracted from the environmental basic disturbance data. The technical effect is reflected in the accurate positioning of risk-related information in the environmental dimension, providing a clear basis for the subsequent construction of environmental correction coefficients.

[0038] By classifying interference types, calculating dynamic coefficients, and integrating them, the environmental risk correlation data output in step S302 is transformed into environmental correction correlation data. First, the environmental risk correlation data is divided into three independent interference types: wind speed correction, temperature correction, and terrain correction, to avoid cross-influence during correction. Then, based on power safety regulations and statistical data on fall accidents caused by environmental factors over the past five years, dynamic coefficient calculation rules are formulated, assigning corresponding coefficients according to the interference intensity gradient (the coefficient range corresponds to no risk to extremely high risk). Next, the single-type correction coefficients for each type of interference are calculated separately, and the maximum value is used to integrate them into a comprehensive environmental correction coefficient (prioritizing the environmental interference with the greatest impact). Finally, the coefficients are validated to ensure consistency with the environmental risk correlation data. This transforms qualitative environmental risk correlation information into quantitative correction coefficients, making environmental interference risks calculable. Taking mountaintop climbing operations as an example, correlation data with high wind speeds and temperatures near the safety limit at the mountaintop can be transformed into corresponding comprehensive environmental correction coefficients. The technical effect is reflected in dynamically adapting to different interference intensities, avoiding the one-size-fits-all problem of fixed coefficients, and ensuring that the correction direction is consistent with the actual environmental risk drivers.

[0039] Using operational process data collected through power operation site process control technology as input, basic operational data is extracted through dimensional segmentation, data filtering, and format standardization. The operational process data originates from the power production management system (stores work instructions and step execution records), the monitoring personnel recorder (collects monitoring personnel on-duty status), the work hour system (records work duration and personnel rotation), and the work permit management system (links monitoring qualifications, etc.). The filtering process focuses on three core dimensions: progress and duration, personnel configuration and monitoring, and step execution. Redundant information such as work plan numbers and work permit submitter contact information is removed, while actual work progress, cumulative duration, personnel configuration ratio, monitoring personnel on-duty status, step execution compliance, and safety measure implementation are retained. Simultaneously, the data is standardized into a standardized format (e.g., monitoring status is coded, step execution is hierarchical), and the completeness of the filtered data is verified, triggering supplementary data collection when core dimensions are missing. This process data, scattered across multiple systems, is transformed into structured basic operational information, laying the data foundation for subsequent focus on process risk-related content. Taking crossarm operation in power distribution network as an example, basic operational data such as current operation progress, cumulative operation time, on-duty monitoring personnel, and compliance of step execution can be filtered out. The technical effect is reflected in removing irrelevant information from the process data, ensuring that the retained content is directly related to the risk of falling from heights, and providing reliable data support for subsequent process risk analysis.

[0040] Based on the basic operational data output in step S304, risk-related data is extracted through risk dimension mapping, key information filtering, and correlation verification. First, a correspondence is established between the basic operational dimensions of the process and the fall risk dimension, clarifying that progress and duration correspond to personnel fatigue risk, personnel allocation and supervision correspond to regulatory gap risk, and step execution corresponds to operational omission risk. Then, based on this correspondence, information in the basic operational data that has a direct causal relationship with fall risk is filtered, and redundant content such as non-safety-related details in the step execution records is removed. Finally, the filtering results are verified for risk correlation to ensure that all retained information directly affects fall safety in high-altitude operations. This approach further focuses on the core risks from the structured basic operational data, avoiding insufficient targeting due to information redundancy when extracting subsequent process correction correlation data. Taking distribution network crossarm operations as an example, correlation data such as the impact of cumulative working time on personnel fatigue, the role of on-duty supervision in ensuring regulation, and the support of step compliance for operational safety can be extracted from the basic operational data. The technical effect is reflected in the accurate positioning of risk-related information in the process dimensions, providing a clear basis for subsequently constructing process correction coefficients.

[0041] By classifying process risk types, calculating dynamic coefficients, and integrating them comprehensively, the process risk-related data output in step S305 is transformed into process correction-related data. First, the process risk-related data is divided into three independent risk types: duration correction, monitoring correction, and procedure correction, to avoid cross-influence during correction. Then, based on power safety regulations and statistical data on fall accidents caused by process deviations over the past five years, dynamic coefficient calculation rules are formulated, assigning corresponding coefficients according to the degree of deviation (the coefficient range corresponds to no risk to extremely high risk). Next, the single-type correction coefficients for each type of risk are calculated separately, and a weighted summation principle (based on accident statistics to determine the weight of each type) is used to integrate them into a comprehensive process correction coefficient (reflecting the cumulative effect of multiple process risks). Finally, the coefficients are validated to ensure consistency with the process risk-related data. This transforms qualitative process risk-related information into quantitative correction coefficients, making process deviation risks calculable. Taking distribution network crossarm operations as an example, correlation data with long cumulative operation time, compliant monitoring, and compliant procedures can be transformed into corresponding comprehensive process correction coefficients. The technical effect is reflected in accurately reflecting the cumulative impact of multiple process risks, avoiding the limitations of single-risk correction, and ensuring that the correction direction is consistent with the actual process risk driving factors.

[0042] Using the environmental correction correlation data output in step S303, the process correction correlation data output in step S306, and the initial risk feature vector output in step S2 as inputs, a calibration risk feature vector is generated through correction weight determination, dimensional correlation matching, and vector integration verification. First, based on the risk contribution ratio of environmental and process factors in fall accidents over the past five years, the correction weights of the environmental correction correlation data and the process correction correlation data are determined to ensure a balanced total weight in the corrected vector. Then, the correlation strength (divided into strong correlation, weak correlation, and no correlation) between each dimension (personnel and equipment dimensions) of the initial vector and the two types of correction coefficients is analyzed. Strongly correlated dimensions directly apply the correction coefficient and corresponding weight to calculate the corrected value; weakly correlated dimensions use a discount coefficient to weaken the correction impact; and no correlation dimensions retain their original values. Next, the corrected values ​​of each dimension are integrated into the calibration risk feature vector in their original order, and the adaptability of the vector to the characteristics of the current work process is verified (e.g., strengthening the correction effect on the physiological dimensions of personnel in crossarm operations). Finally, the rationality of the vector values ​​is verified. The quantification of environmental dynamic interference and process deviation risks is integrated into the initial vector to generate a comprehensive risk representation that fits the actual working conditions. Taking crossarm operation in distribution networks as an example, a calibration vector reflecting the comprehensive risk of four dimensions can be obtained by adjusting strongly correlated dimensions such as the physiological adaptability of personnel and the effectiveness of equipment functions in the initial vector. The technical effect is reflected in the comprehensiveness of risk characterization and the adaptability of links, providing accurate input for subsequent full-process risk assessment.

[0043] In one embodiment, the step of obtaining the risk transmission path between each work stage and the cross-stage correlation strength between adjacent work stages based on the real-time data, and generating a high-altitude operation fall risk correlation map by combining the risk transmission path and the cross-stage correlation strength, includes: S401. Obtain basic time-series data of the work process based on the real-time data; S402. Obtain the link connection data based on the basic time sequence data of the work process; S403. Obtain initial risk transmission path data based on the link connection data; S404. Obtain basic data for risk interaction in the process based on real-time data; S405. Obtain raw data on the intensity of risk association in each stage based on the basic data of risk interaction in each stage; S406. Obtain cross-stage correlation strength adaptation data based on the original data of stage risk correlation strength; S407. Generate a high-altitude operation fall risk correlation diagram based on the initial risk transmission path data and cross-link correlation strength adaptation data.

[0044] As described in steps S401-S407 above, this invention uses multi-source real-time data from power high-altitude operations as input. Through time-series dimension definition, data filtering, format standardization, and integrity verification, it extracts basic time-series data for the work process. Real-time data originates from the power production management system (stores work sequence plans and start / stop records), the worker positioning system (collects location-time trajectories for each stage), the monitoring personnel recorder (extracts voice timestamps for stage transitions), and the work time statistics system (records stage duration and time-series deviations). The filtering process focuses on three dimensions: stage time-series nodes (start / stop time, time-series deviation identifiers), stage transition sequence (transition intervals, triggering conditions), and overall time-series progress (cumulative time consumption, time-series compliance). Redundant information such as work plan numbers and personnel names is removed. The time format, time-series deviation identifiers, and transition triggering conditions are standardized into a uniform format. The filtered data is verified, and supplementary data collection is triggered when core dimensions are missing. This transforms unstructured time-series data scattered across multiple systems into structured basic time-series information for the work process, laying the foundation for subsequent extraction of stage-series connection data. Taking the maintenance of transmission line towers as an example, data such as the actual start and stop times, connection triggering conditions, and timing deviations of each link can be screened out. The technical effect is reflected in the realization of time-series data structuring and deviation explicitness, avoiding interference from redundant or formatted data in subsequent path extraction.

[0045] Based on the basic time-series data of the work process output in step S401, the linkage data of each stage is extracted through time-series-connection mapping, key information filtering, and correlation verification. First, the correspondence between the dimensions of the basic time-series data and the dimensions of stage connections is established, clarifying the sequence of stages (start and stop times), the tightness of connections (intervals), the safety assurance (trigger conditions), and the risks associated with time-series deviations. Then, information directly related to risk transmission is filtered, eliminating redundant details such as planned stage duration and specific operational steps. Finally, the risk transmission correlation of the filtered results is verified to ensure that the retained information directly affects the risk transmission effect between stages. Focusing on the core risks of stage connections from structured time-series data avoids subsequent path extraction that only considers the sequence of events while ignoring risk correlations. Taking the connection between pole climbing and crossarm operations in transmission line tower maintenance as an example, the sequence of the two stages, whether the connection interval meets safety inspection requirements, whether the trigger conditions include safety confirmation, and whether there are potential risks from exceeding the time limit for previous pole climbing can be extracted. The technical effect is reflected in the accurate positioning of connection risk elements, providing a risk-driven basis for the initial risk transmission path data extraction.

[0046] By structuring the connection relationships, defining path directions, integrating path sequences, supplementing attributes, and verifying logic, the data related to the connection of links is transformed into initial risk transmission path data. First, a unique identifier is assigned to each work link, transforming the connection sequence into a link pair structure of "preceding link - subsequent link". Second, directed edge attributes are added to the link pairs to clarify the risk transmission direction. Then, directed link pairs are linked together according to the actual work sequence to form a path sequence, and based on the safety guarantees and potential risks in the connection data, path attributes such as low-risk transmission and high-risk risks are added to each directed edge. Finally, the path sequence is verified to ensure it conforms to the safety logic of high-altitude power operations, eliminating reverse or closed-loop paths. Qualitative connection information is transformed into structured, directional initial paths with risk attributes, providing a path framework for subsequent connection graph construction. Taking transmission line tower maintenance as an example, a path sequence can be formed: climbing the pole towards crossarm operation, crossarm operation towards conductor / ground wire operation, and conductor / ground wire operation towards pole descent, with risk attributes marked for each path. The technical effect is reflected in achieving a visual foundation for path visualization and explicit risk characteristics, ensuring that the path is consistent with the actual work safety logic.

[0047] Using multi-source real-time data from high-altitude power operations as input, this study extracts fundamental data on risk interactions at each stage through risk interaction dimension definition, data filtering, format standardization, and integrity verification. Real-time data originates from equipment status sensors (collecting the impact of preceding equipment defects on subsequent stages), computer vision systems (recording the impact of preceding personnel operations on subsequent stages), risk event recording systems (reporting risk impacts between stages), and historical accident correlation databases (assisting in identifying interaction patterns). The filtering process focuses on three dimensions: interaction subjects (affected stages, affected stages, and risk factor types), interaction behavior (the specific process of risk transmission), and interaction impact (the results and duration on affected stages). Redundant information such as equipment production batches and personnel training schools is removed. The description formats for interaction subjects, behaviors, and impacts are standardized, and data integrity is verified. Supplementary data collection is triggered when core dimensions are missing. This transforms implicit inter-stage risk interaction processes into structured fundamental information, providing interactive data support for subsequent correlation strength calculations. Taking the impact of foot strap wear on crossarm operation during the pole climbing process in transmission line tower maintenance as an example, we can extract basic data such as the influencing link being pole climbing, the affected link being crossarm operation, the risk factor being equipment factor, and the degree and duration of the impact caused by the wear of foot straps leading to increased personnel posture adjustment during crossarm operation. The technical effect is reflected in making implicit interaction explicit and standardizing data, which facilitates subsequent intensity quantification calculation.

[0048] By classifying interaction types, formulating intensity calculation rules, calculating raw intensity, and verifying validity, the basic data of risk interactions in each stage are transformed into raw data of risk-related intensity in each stage. First, risk interactions are categorized into three independent types based on factor type: personnel, equipment, and environment. Then, based on power safety regulations and historical accident data, calculation rules are established for the base value corresponding to the degree of impact and the correction coefficients corresponding to the factor type and duration. The raw intensity is obtained by multiplying the base value by the two correction coefficients. Finally, the consistency between the raw intensity and the basic interaction data is verified. Qualitative or semi-quantitative interaction information is transformed into quantitative raw intensity in the 0-1 range, providing a calculable basis for subsequent intensity adaptation. Taking the interaction of pole climbing clip wear on crossarm operations during transmission line tower maintenance as an example, if the interaction is an equipment factor, has a moderate impact, and continuously covers the entire crossarm operation, the corresponding raw intensity can be calculated. The technical effect is reflected in achieving objective quantification of intensity and explicit disclosure of risk differences, avoiding the subjective bias of traditional experience-based assignment, and ensuring a unified standard for intensity calculation for different interaction events.

[0049] By employing path-strength matching, dynamic correction, format standardization, and adaptability verification, the original data on the risk-related strength of each stage is transformed into cross-stage related strength adaptive data. First, based on the stage pairs in the initial risk transmission path data, the original strength is matched with the corresponding path; paths without a matching original strength are assigned a default base strength. Second, considering the risk characteristics (such as high hazard, low risk) in the path attributes, the matched strength is adjusted using a correction coefficient. Then, the data is integrated according to "path stage pair - corrected strength - correction reason". Finally, it is verified whether the corrected strength contradicts the path attributes. Binding independent original strength data to the risk characteristics of the initial path ensures that the strength reflects the superposition of the path's own safety status and the interaction effects of each stage. Taking the path for climbing a pole to crossarm operation in transmission line tower maintenance as an example, if the original strength match is medium and the path attribute is high hazard, an adaptive strength can be obtained through a correction coefficient, and the correction reason can be marked. The technical effect is reflected in achieving precise path-strength binding and transparency in the correction process, ensuring that the strength dynamically reflects the differences in path risk and facilitating subsequent risk tracing.

[0050] By defining the graph structure, integrating data, verifying the topology, and outputting a structured graph, an association graph of fall risk in high-altitude operations is generated by merging initial risk transmission path data with cross-stage correlation strength adaptation data. First, the graph is defined with three main elements: nodes (operation stages, including identifier, name, and risk attribute), directed edges (risk transmission paths, including start point, end point, and path attribute), and edge weights (cross-stage correlation strength, including corrected strength and correction reason). Then, the two types of data are integrated to fill the graph elements. Next, a depth-first search algorithm is used to verify the graph's topology, eliminating closed-loop paths and isolated nodes. Finally, the graph is output in a structured format with node tables, edge tables, and edge weight tables. This integrates scattered path and intensity data into a standardized, visualized, and computable global graph model, clearly presenting the risk transmission relationships throughout the entire process. Taking transmission line tower maintenance operations as an example, the output association graph can intuitively display the global structure with each operation stage as a node, the risk transmission direction as a directed edge, and the corrected strength as the edge weight. The technical effect is reflected in achieving global risk visualization and model callability, providing a global risk linkage framework for subsequent calculation of the overall fall risk value for operations.

[0051] In one embodiment, the step of obtaining the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram includes: S501. Obtain the link risk dimension data and graph structure core data based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. S502. Obtain local risk characterization data of the process based on the process risk dimension data and graph structure core data; S503. Obtain risk transmission impact data based on local risk characterization data and core graph structure data. S504. Obtain comprehensive risk data for each stage based on risk transmission impact data and local risk characterization data of each stage; S505. Based on the node relationships in the comprehensive risk data of each process and the core data of the graph structure, obtain the risk linkage data of the entire process; S506. Obtain overall risk aggregation basic data based on the full-process risk linkage data; S507. Based on the overall risk aggregation basic data, key risk information is structured and integrated to obtain the overall fall risk value of the operation.

[0052] As described in steps S501-S507 above, this invention takes the calibration risk feature vector output in step S3 and the high-altitude operation fall risk association diagram output in step S4 as inputs. Through vector dimension analysis, association diagram structure decomposition, data format unification, and integrity verification, it extracts the risk dimension data of each stage and the core data of the graph structure. The calibration risk feature vector contains the personnel and equipment dimension risk quantification values ​​of each operation stage. The high-altitude operation fall risk association diagram contains a node (operation stage), directed edge (risk transmission path), and edge weight (cross-stage association strength) structure. In the process, step S501: First, the 8 core dimensions of the calibration vector (4 for personnel and 4 for equipment) are decomposed, and the identifier, risk type, and quantification value range of each dimension are extracted and bound to the corresponding operation stage; then, the association diagram is decomposed into node core data (identifier, name, risk attribute), directed edge core data (start point, end point, path attribute), and edge weight core data (value, correction reason), and the association between edges and edge weights is established; finally, the two types of data are unified into a "key-value" format, and it is verified whether it covers all stages and paths. If missing, supplementary data is collected. By breaking down two types of complex input data into independently processable "local risk units" and "related structural units," the computational logic can be kept clear. Taking transmission line tower maintenance as an example, data on the personnel qualification compliance dimension of the pole climbing process and the edge weight data of the pole climbing leading to the crossarm operation in the correlation diagram can be extracted. The technical effect is reflected in the clear data decomposition, which provides an accurate data foundation for subsequent local risk characterization and transmission impact calculation, and avoids confusion of risk dimensions.

[0053] Step S502: Obtain local risk characterization data for each stage based on the stage risk dimension data and graph structure core data. This step generates local risk characterization data for each stage based on the stage risk dimension data and graph structure core data through node-dimension matching, risk dimension screening, local risk quantification, and characterization data integration. First, match the work stage nodes with the corresponding risk dimensions according to "node identifier = stage identifier" to ensure that nodes such as pole climbing and crossarm operations only match their own dimension data; then, screen the core risk dimensions according to power safety regulations and eliminate basic compliance dimensions such as training compliance; then calculate the risk contribution of each dimension and sum them to obtain the local risk value of the stage; finally, integrate the node identifier, core dimension list, local risk value, and out-of-standard dimension identifier to form structured data. Aggregate the scattered dimension data into quantifiable local risk characterization and clarify the risk level of each stage. Taking the pole climbing stage of transmission line tower maintenance as an example, its local risk value can be calculated, and the existence of high-risk dimensions can be identified. The technical effect is reflected in the accurate quantification of local risks, the direct comparison of risks in different stages, the focus on core risks, and the avoidance of basic dimensions diluting key risks.

[0054] Step S503: Obtain risk transmission impact data based on local risk characterization data and core graph structure data. This step quantifies the risk transmission impact of preceding stages on subsequent stages through transmission path location, correlation strength matching, transmission impact calculation, and transmission data integration. First, the correlation graph is traversed to locate transmission paths such as pole climbing leading to crossarm operation and crossarm operation leading to conductor / ground wire operation, and path attributes are marked. Then, the edge weight values ​​corresponding to each path are matched; if no matching edge weight is found, a default value is assigned. Subsequently, based on the local risk value of the preceding stage, path edge weight, path attribute correction coefficient, and transmission attenuation coefficient, the transmission impact is decomposed into each risk dimension, integrating path identifiers, transmission impact values, and dimension contributions to form structured data. This transforms the implicit cross-stage risk transmission effect into quantitative data, providing a transmission impact benchmark for subsequent comprehensive risk calculation. Taking transmission line tower maintenance operations as an example, the transmission impact of the pole climbing stage on crossarm operation can be accurately quantified. The technical effect is reflected in the dynamic nature of the transmission impact and the explicit manifestation of differences in transmission intensity across different paths, avoiding the limitations of traditional fixed coefficients. Simultaneously, the decomposition of dimension contributions facilitates subsequent targeted management.

[0055] Step S504: Obtain comprehensive risk data for each stage based on risk transmission impact data and local risk characterization data. This step integrates local risks of subsequent stages with transmission impacts of preceding stages through stage-transmission impact matching, transmission impact superposition, duplicate risk elimination, and comprehensive risk integration. First, subsequent stages are matched with their corresponding transmission impacts based on the "endpoint stage identifier equals node identifier." For multiple preceding stages, they are weighted and superimposed using a direct preceding stage weight of 0.7 and an indirect preceding stage weight of 0.3. Then, the contribution of transmission impacts and local risks in the same dimension is compared, and the maximum value is used to eliminate duplicate risks (such as the equipment adaptability contribution of pole climbing transmission and the equipment adaptability contribution of crossarm operation, retaining the larger value). Subsequently, the comprehensive contribution of each dimension is summed to obtain the comprehensive risk value of the stage. Finally, the stage identifier, total transmission impact value, comprehensive risk value, and high-risk dimensions are integrated to form structured data. This generates a complete risk characterization containing "local risks and transmission impacts," avoiding risk underestimation caused by isolated calculations. Taking the crossarm operation of transmission line tower maintenance as an example, its comprehensive risk value includes its own local risks and the impact of pole climbing and transmission. The technical effect is reflected in the complete risk characterization, rationalization of repeated risks, strong adaptability to multiple preceding steps, and ensuring that the risk value is consistent with the actual working conditions.

[0056] Step S505: Obtain full-process risk linkage data based on the node relationships in the comprehensive risk data of each stage and the core data of the graph structure. This step generates full-process risk linkage data based on the node relationships in the comprehensive risk data of each stage and the core data of the graph structure through node relationship analysis, comprehensive risk chaining, key node identification, and linkage data integration. First, analyze the node relationships and sort out the sequence of work stages according to the directed edges of the association graph (e.g., pole climbing, crossarm operation, conductor grounding wire operation, pole lowering), clarifying the preceding and following stages of each node; then, chain the comprehensive risk values ​​of each stage in this order and mark the proportion of transmission impact; subsequently, based on the comprehensive risk value, node attributes (initial or intermediate core stages are key nodes), and the proportion of transmission impact, identify key nodes; finally, integrate the node association chain, risk sequence, and key node identifiers to form structured data. This transforms the scattered stage risks into sequence data reflecting the linkage effect of the entire process, realizing the visualization of risk transmission trends. Taking the maintenance of transmission line towers as an example, the entire process trend of risk transmission from pole climbing to crossarm operation and then attenuation to conductor and ground wire operation can be clearly presented. The technical effect is reflected in the visibility of key nodes and the traceability of risk sources, providing a global view for subsequent focus on core risks.

[0057] Step S506: Obtaining Overall Risk Aggregation Base Data Based on Full-Process Risk Linkage Data. This step focuses on core risk information from the full-process risk linkage data through key node screening, core risk extraction, redundant data removal, and basic data integration. First, based on key node identifiers, key nodes such as pole climbing and crossarm operations are screened, while non-key nodes such as conductor and ground wire operations and pole removal are removed. Then, the core risk information of key nodes (identifier, comprehensive risk value, proportion of transmission impact, and edge weight of associated paths) is extracted, and redundant content such as specific operation steps is removed. Finally, the core risks of key nodes and the transmission strength of key paths (such as the edge weight of pole climbing leading to crossarm operations) are integrated to form structured overall risk aggregation base data. This simplifies risk information, focusing on key nodes and paths that contribute significantly to the overall process risk, and avoiding interference from non-key information in the aggregation calculation. Taking transmission line tower maintenance operations as an example, only the core risks and transmission strength data of pole climbing and crossarm operations are retained. The technical effect is reflected in the accuracy of the aggregation basis, reduced calculation complexity, and preservation of the risk linkage relationship between key nodes, ensuring that subsequent aggregation results reflect core risks.

[0058] Step S507: The first is the core risk data of key nodes, which is the node data whose contribution to the overall process risk exceeds the safety ratio. For example, the comprehensive risk value of pole climbing (initial key node) and crossarm operation (intermediate core node) is represented by the superposition result of the node's own local risk and the impact of the preceding links. It is within a specific reasonable range and can accurately reflect the current risk status of the node. The second is the key transmission path data, which includes the edge weights between key nodes (the strength of cross-link association, within a specific reasonable range) and the corresponding transmission impact value (the quantitative result of the risk transmission from the preceding node to the subsequent node, within a specific reasonable range), which can reflect the intensity of risk transmission between links. The weighting rules are based on statistical data of falls from heights during power operations over the past five years. Initial critical nodes, which directly determine the safety foundation of subsequent stages, are assigned higher weights than intermediate core nodes. The weight of a critical transmission path is the product of the corresponding path edge weight and the total weight of the reserved transmission impact, ensuring that the sum of all weights is 1. This provides a balanced weighting basis for subsequent quantitative calculations and avoids any one factor excessively dominating the risk value. Initial critical nodes and intermediate core nodes have different risk contribution roles in the entire process, and are assigned different weights. The weight of initial critical nodes is higher than that of intermediate core nodes. This is because the initial critical node is the starting point of the operation, and its risk status directly affects all subsequent stages. Although intermediate core nodes have a significant risk contribution, they are greatly affected by the initial stage, and the weighting needs to reflect this difference.

[0059] This calculation method quantifies and weights the risk of each key node according to its actual contribution, avoiding risk value deviations caused by using the same weight for nodes of different importance. The overall temporary risk value is calculated as follows: "Overall temporary risk value equals the sum of the node's comprehensive risk value × node weights, plus the product of the transmitted impact value and the transmitted weight." Substituting the key nodes and transmitted impact values ​​from the overall risk aggregation base data, a temporary value is obtained. Then, a min-max normalization algorithm is used to map the temporary value to the [0,100] interval. The formula for calculating the overall fall risk value is: ; in, This indicates the overall fall risk value of the operation. Indicates a temporary value. Represents the minimum temporary value. This represents the maximum temporary value, the minimum temporary value when all stages of risk are safe, and the maximum temporary value when all stages of risk are extremely high. Finally, the standardized value is verified, and the results are output along with the calculation basis. Core risk information is aggregated into standardized values ​​that can be directly used for risk assessment, providing a benchmark for subsequent threshold judgments. Taking transmission line tower maintenance as an example, the standardized overall risk value can intuitively reflect the risk level of the entire process. The technical effect is reflected in the intuitiveness of risk assessment, the transparency and traceability of the calculation process, and ensuring that it matches the actual accident risk level.

[0060] In one embodiment, the step of generating graded early warning information including risk level and control strategy based on the overall fall risk value includes: S601. Obtain basic risk characterization data based on the overall fall risk value data; S602. Obtain risk level determination data based on risk basic characterization data; S603. Obtain risk warning level data based on risk level determination data; S604. Obtain the association data between risk sources and transmission paths based on risk warning level data and previously obtained high-altitude operation fall risk association diagrams. S605. Obtain basic data for adaptive control strategies based on the correlation data between risk sources and transmission paths; S606. Obtain structured component data of graded early warning information based on the basic data of the adaptation and control strategy and the risk warning level data. S607. Obtain the risk level and control strategy corresponding to each component data based on the structured component data of the graded early warning information.

[0061] As described in steps S601-S607 above, this invention uses the overall fall risk value data output in step S507 as input, and generates basic risk characterization data through risk value dimension expansion, trend analysis, scenario association, and data integration. The overall fall risk value data includes the overall fall risk value of the operation, the trend of risk value changes, the proportion of risk contribution, and the operation scenario identifier. First, four core dimensions are extracted, and redundant information such as temporary values ​​from aggregation calculations is removed. Then, based on the risk value sampling data from the past ten minutes, the fluctuation amplitude is calculated, and the trend types of stable, slight fluctuations, or severe fluctuations are marked. The rationality of the trend is analyzed in conjunction with the characteristics of the operation scenario (such as high towers, susceptibility to gusts). Subsequently, the main sources of risk contribution (nodes or paths with a contribution ratio exceeding 30%) are identified, clarifying the core risk sources. Finally, the data is integrated according to the logic of "risk value-trend-contribution-scenario" to form a structured characterization. Furthermore, the single quantitative risk value is expanded into a multi-dimensional description, overcoming the limitation that numerical values ​​alone cannot reflect the stability, source, and scenario characteristics of risk. Taking pole climbing operations on transmission lines as an example, it can clearly identify the range of risk values, whether the changes are stable, and whether the risk mainly comes from the pole climbing process or the transmission path. The technical effect is reflected in the comprehensiveness of risk information, providing multi-dimensional basis for subsequent level determination and avoiding one-sided judgments caused by a single value.

[0062] Based on risk-based characterization data, a pre-defined rule library for determining the risk level of high-altitude power operations is invoked. Through scenario threshold matching, trend correction, contribution source correction, and data integration, risk level determination data is generated. The rule library contains basic risk thresholds, risk trend correction coefficients, and contribution source correction coefficients for different operation scenarios. The process involves first matching the corresponding basic threshold according to the operation scenario identifier; then adjusting the threshold based on the risk value change trend; subsequently, further correction based on the main contribution source type; and finally, integrating the basic thresholds, correction process, actual determination thresholds, and supplementary explanations to form a structured basis. This transforms general determination rules into personalized standards adapted to the current scenario, trend, and contribution source, avoiding the limitations of traditional fixed thresholds. Taking power transmission line pole climbing operations as an example, if the main contribution source is equipment factors and the trend is stable, the actual determination threshold can be accurately calculated. The technical effect is reflected in the scenario-based determination basis and the transparent correction logic, ensuring that subsequent level determinations are highly matched with actual risk characteristics.

[0063] The system utilizes a risk warning level standard library for high-altitude power operations. Based on risk level determination data, it generates risk warning level data through risk value range matching, level identifier assignment, hazard description association, and level data integration. The standard library includes color identifiers, level names, accident probabilities, possible consequences, and response priorities for four warning levels. The process involves first matching the overall fall risk value of the operation with the actual judgment threshold to preliminarily determine the risk level; then matching the corresponding level identifier; subsequently, associating the accident probability and possible consequences for that level; and finally, combining the supplementary explanations in the risk level determination criteria to supplement hazard details and response priorities, integrating them into structured data. This transforms abstract threshold ranges into concrete warning levels, clearly defining the degree of risk hazard and the urgency of the response. Taking power transmission line pole climbing operations as an example, the warning level can be clearly defined as Level III (blue), corresponding to a medium response priority. The technical effect is reflected in the intuitive level and precise hazard description; on-site personnel can quickly identify risks through color and name and respond in an orderly manner according to priority.

[0064] Using risk warning level data and the high-altitude operation fall risk correlation diagram output from step S4 as input, this process generates risk source and transmission path correlation data through node location, risk factor matching, transmission path tracing, and correlation data integration. The correlation diagram includes operation link nodes, risk transmission paths, and edge weights. The process involves first locating key nodes in the correlation diagram based on the supplementary explanations of the risk warning level data; then selecting specific factors consistent with the contributing source type from the core risk factors associated with the nodes; subsequently tracing the output transmission path of key nodes and determining the risk transmission channel based on edge weights; finally, integrating key nodes, risk source factors, risk source locations, main transmission paths, and hazard descriptions to form structured data. Furthermore, the macro-level risk is broken down into micro-level risk sources and transmission paths, achieving precise risk tracing. Taking power transmission line pole climbing operations as an example, the risk source can be clearly identified as insufficient foot strap adaptability in the pole climbing stage, and the transmission path is from pole climbing to crossarm operation. The technical effect is reflected in targeted tracing and visualized transmission, avoiding blind subsequent control measures.

[0065] The system invokes a pre-set database of risk control strategies for high-altitude power operations. Based on the correlation data between risk sources and transmission paths, it generates basic data for adaptive control strategies through risk source strategy matching, intensity adjustment, path blocking strategy supplementation, and basic data integration. The database contains basic strategies corresponding to different risk source factors, matching rules for warning levels and strategy intensity, and transmission path blocking measures. The process involves first matching basic strategies according to risk source factors; then adjusting the strategy execution intensity based on the risk warning level; subsequently supplementing blocking measures for major transmission paths; and finally clarifying the responsible person and completion deadline for each strategy, integrating the data into structured data. The risk tracing results are then transformed into actionable control strategies, ensuring that risk management moves from "knowing the problem" to "solving the problem." Taking power transmission line pole climbing operations as an example, it can be clearly stated that the person in charge of the operation will lead the inspection of the foot catches, which must be completed within ten minutes. The technical effect is reflected in the specific and implementable strategy, clear responsibility and deadlines, and avoidance of shirking responsibility and execution deviations.

[0066] By defining component dimensions, populating component content, verifying component consistency, and integrating structured data, the basic data for adapting control strategies and risk warning level data are decomposed into standardized, hierarchical, structured component data for early warning information. Three core components are defined: risk level component, risk source and path component, and control strategy component. During the process, content is extracted from the corresponding input data to populate each component, and the logical consistency between components is verified (e.g., high-risk levels require matching shutdown strategies). Components are integrated in the order of "risk level, risk source path, control strategy." Dispersed risk information is decomposed into standardized, combinable component units to adapt to the needs of different push channels. Taking power transmission line pole climbing operations as an example, the component data includes a blue Level III warning, foot clip compatibility risk sources, and foot clip inspection strategies. The technical effect is reflected in clear information logic, independent and combinable components, the ability for audible and visual alarms to push only the level component, and the ability for mobile devices to push the complete component, balancing information conciseness and detail.

[0067] Through component logic integration, content verification, format standardization, and final information output, the structured component data of hierarchical early warning information is integrated into complete hierarchical early warning information. During this process, the text is integrated according to the cognitive logic of "risk level components, risk source and path components, and control strategy components" to ensure coherent and consistent statements; the completeness and logical consistency of core information are verified; the standard early warning format of the power industry is adopted, including the early warning title, level, risk source tracing, control requirements, effective time, and feedback requirements; finally, it is simultaneously pushed to on-site audible and visual alarms, mobile apps for operators, and the power safety production management and control platform. The component data is transformed into a final early warning instruction that is "pushable, executable, and traceable," forming a closed loop for risk management. Taking transmission line pole climbing operations as an example, the early warning information clearly defines the blue Level III warning, the risk source of the foot strap, and the inspection requirements. The technical effect is reflected in the completeness and omission of information, multi-channel coverage, and closed-loop feedback mechanism, allowing on-site personnel to respond efficiently and provide feedback, balancing operational safety and efficiency.

[0068] like Figure 2 As shown, the present invention also provides a dynamic assessment and early warning system for fall risks during high-altitude power operations, comprising: Real-time data acquisition module 1 is used to acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple work stages during high-altitude power operations. The initial risk feature vector acquisition module 2 is used to extract the risk factors corresponding to each operation step based on the real-time personnel status data and equipment status data, and to obtain the initial risk feature vector of each operation step based on the risk factors. The calibration risk feature vector acquisition module 3 is used to dynamically correct the initial risk feature vector based on the environmental data and work process data to obtain the calibration risk feature vector corresponding to each work step. The fall risk association diagram generation module 4 is used to obtain the risk transmission path between each operation link and the cross-link association strength between adjacent operation links based on the real-time data, and generate a high-altitude operation fall risk association diagram by combining the risk transmission path and the cross-link association strength. The fall risk value acquisition module 5 is used to acquire the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. Module 6 is used to determine whether the overall fall risk value exceeds a preset risk threshold. If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

[0069] In one embodiment, the initial risk feature vector acquisition module 2 includes: A personnel basic attribute data acquisition unit is used to acquire personnel basic attribute data based on the real-time personnel status data. The associated data acquisition unit is used to acquire personnel risk associated data based on the personnel basic attribute data; A personnel risk factor set acquisition unit is used to acquire a personnel risk factor set based on the personnel risk association data. The device basic attribute data acquisition unit is used to acquire device basic attribute data based on the real-time device status data. The equipment risk association data acquisition unit is used to acquire equipment risk association data based on the equipment basic attribute data; The equipment risk factor set acquisition unit is used to acquire the equipment risk factor set based on the equipment risk association data. The initial risk feature vector acquisition unit is used to acquire the initial risk feature vector of each operation link based on the personnel risk factor set and the equipment risk factor set.

[0070] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described dynamic assessment method for fall risk in high-altitude power operations.

[0071] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described dynamic assessment method for fall risk in high-altitude power operations.

[0072] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0073] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0074] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent results or equivalent process transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for dynamic assessment and early warning of fall risk during high-altitude power operations, characterized in that, include: Acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple operational stages during high-altitude power operations; Based on the real-time personnel status data and equipment status data, extract the risk factors corresponding to each work step, and obtain the initial risk feature vector of each work step based on the risk factors. Based on the environmental data and work process data, the initial risk feature vector is dynamically corrected to obtain the calibration risk feature vector corresponding to each work step. Based on the real-time data, the risk transmission path between each work stage and the cross-stage correlation strength between adjacent work stages are obtained. A high-altitude operation fall risk correlation diagram is generated by combining the risk transmission path and the cross-stage correlation strength. The overall fall risk value of the operation is obtained based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. Determine whether the overall fall risk value exceeds a preset risk threshold; If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

2. The method for dynamic assessment and early warning of fall risk in high-altitude power operations according to claim 1, characterized in that, The step of extracting risk factors corresponding to each work step based on the real-time personnel status data and equipment status data, and obtaining the initial risk feature vector of each work step based on the risk factors, includes: Based on the real-time personnel status data, obtain the basic attribute data of the personnel; Based on the aforementioned basic personnel attribute data, obtain personnel risk correlation data; Obtain a set of personnel risk factors based on the aforementioned personnel risk correlation data; Obtain basic device attribute data based on the real-time device status data; Obtain equipment risk association data based on the aforementioned basic equipment attribute data; Obtain the set of equipment risk factors based on the equipment risk correlation data; The initial risk feature vectors for each operational stage are obtained based on the personnel risk factor set and the equipment risk factor set.

3. The method for dynamic assessment and early warning of fall risk in high-altitude power operations according to claim 1, characterized in that, The step of dynamically correcting the initial risk feature vector based on the environmental data and work process data to obtain the calibrated risk feature vector corresponding to each work step includes: Obtain basic environmental disturbance data based on environmental data; Environmental risk correlation data are obtained based on basic environmental disturbance data; Obtain environmental remediation correlation data based on environmental risk correlation data; Obtain basic operational data of the process based on the workflow data; Obtain process risk-related data based on the basic process operation data; Obtain process correction-related data based on process risk-related data; Based on the environmental correction correlation data and process correction correlation data, the initial risk feature vector is adjusted for dimension adaptation to obtain the calibration risk feature vector corresponding to each operation link.

4. The method for dynamic assessment and early warning of fall risk in high-altitude power operations according to claim 1, characterized in that, The step of obtaining the risk transmission path between each work stage and the cross-stage correlation strength between adjacent work stages based on the real-time data, and generating a high-altitude operation fall risk correlation map by combining the risk transmission path and cross-stage correlation strength, includes: Based on the real-time data, obtain the basic time sequence data of the work process; Based on the time sequence data of the described work process, obtain the link-related data of each stage; Initial risk transmission path data is obtained based on the data linking the various stages; Based on real-time data acquisition, risk interaction data is obtained in each stage; Obtain raw data on the strength of risk correlation in each stage based on the basic data on risk interaction in each stage; Obtain cross-stage correlation strength adaptation data based on the original data of stage risk correlation strength; A fall risk correlation diagram for high-altitude operations is generated based on the initial risk transmission path data and the cross-stage correlation strength adaptation data.

5. The method for dynamic assessment and early warning of fall risk in high-altitude power operations according to claim 1, characterized in that, The step of obtaining the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram includes: Based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram, obtain the risk dimension data of each link and the core data of the graph structure. Based on the aforementioned risk dimension data and graph structure core data, obtain local risk characterization data for each stage; Risk transmission impact data are obtained based on local risk characterization data and core graph structure data. Comprehensive risk data for each stage is obtained based on risk transmission impact data and local risk characterization data for each stage. Based on the node relationships in the comprehensive risk data of each process and the core data of the graph structure, obtain the risk linkage data of the entire process; Obtain overall risk aggregation base data based on full-process risk linkage data; Based on the overall risk aggregation data, key risk information is structured and integrated to obtain the overall fall risk value for the operation.

6. The method for dynamic assessment and early warning of fall risk in high-altitude power operations according to claim 1, characterized in that, The step of generating graded early warning information including risk level and control strategy based on the overall fall risk value includes: Based on the overall fall risk value data, obtain basic risk characterization data; Data for determining risk levels is obtained based on basic risk characterization data; Risk warning level data is obtained based on the risk level determination criteria. Based on the risk warning level data and the previously obtained high-altitude operation fall risk correlation map, obtain the risk source and transmission path correlation data; Based on the correlation data between risk sources and transmission paths, obtain basic data for appropriate control strategies; Based on the basic data of the adaptive control strategy and the risk warning level data, obtain the structured component data of the graded early warning information; Based on the structured component data of the graded early warning information, the risk level and control strategy corresponding to each component data are obtained from the graded early warning information.

7. A dynamic assessment and early warning system for fall risk during high-altitude power operations, characterized in that, include: The real-time data acquisition module is used to acquire real-time personnel status data, equipment status data, environmental data, and work process data for multiple work stages during high-altitude power operations. The initial risk feature vector acquisition module is used to extract the risk factors corresponding to each operation step based on the real-time personnel status data and equipment status data, and to obtain the initial risk feature vector of each operation step based on the risk factors. The calibration risk feature vector acquisition module is used to dynamically correct the initial risk feature vector based on the environmental data and work process data to obtain the calibration risk feature vector corresponding to each work step. The fall risk association diagram generation module is used to obtain the risk transmission path between each operation stage and the cross-stage association strength between adjacent operation stages based on the real-time data, and generate a high-altitude operation fall risk association diagram by combining the risk transmission path and the cross-stage association strength. The fall risk value acquisition module is used to acquire the overall fall risk value of the operation based on the calibration risk feature vector and the high-altitude operation fall risk correlation diagram. The judgment module is used to determine whether the overall fall risk value exceeds a preset risk threshold; If the overall fall risk value exceeds a preset risk threshold, a graded early warning information containing risk level and control strategy is generated based on the overall fall risk value, and an early warning operation is performed based on the graded early warning information.

8. A dynamic assessment and early warning system for fall risk in high-altitude power operations according to claim 7, characterized in that, The initial risk feature vector acquisition module includes: A personnel basic attribute data acquisition unit is used to acquire personnel basic attribute data based on the real-time personnel status data. The associated data acquisition unit is used to acquire personnel risk associated data based on the personnel basic attribute data; A personnel risk factor set acquisition unit is used to acquire a personnel risk factor set based on the personnel risk association data. The device basic attribute data acquisition unit is used to acquire device basic attribute data based on the real-time device status data. The equipment risk association data acquisition unit is used to acquire equipment risk association data based on the equipment basic attribute data; The equipment risk factor set acquisition unit is used to acquire the equipment risk factor set based on the equipment risk association data. The initial risk feature vector acquisition unit is used to acquire the initial risk feature vector of each operation link based on the personnel risk factor set and the equipment risk factor set.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images