Intelligent safety protection system and method for electric power operation

Abstract

The invention discloses an intelligent safety protection system and method for electric power operation. The system comprises a multi-source fusion positioning and three-dimensional modeling module, a safety state perception and risk pre-judgment module, a hierarchical intervention control module, an edge calculation and digital twinning module, a 5G slice communication transmission module, a multi-scene adaptation module and a system data management and optimization module. The UWB, Beidou and visual reconstruction multi-source fusion positioning technology is adopted, dynamic three-dimensional modeling is combined for positioning deviation correction, the positioning limitation of a single technology in a complex environment is effectively solved, high-precision positioning is achieved, and accurate position support is provided for safety protection. According to the method, infrared thermal imaging, millimeter wave radar and a YOLOv8 optimization algorithm are fused, dominant safety violation behaviors can be identified, hidden misoperation risks can be pre-judged based on data such as equipment temperature and operation postures, and full-dimension safety state sensing is achieved.

Description

Technical Field

[0001] This invention relates to the field of power operation safety protection technology, specifically to an intelligent safety protection system and method for power operations. Background Technology

[0002] With the rapid development of the power industry, complex scenarios such as high-altitude operations on transmission towers and operations in confined spaces like underground substations are becoming increasingly common, placing higher demands on the precision, comprehensiveness, and real-time nature of safety protection for these operations. Currently, power operation safety protection technologies face several bottlenecks: In terms of positioning, existing protection systems mostly use a single GPS or UWB positioning method. In complex environments such as high-altitude obstruction and underground signal jamming, the positioning accuracy is insufficient, making it impossible to accurately capture the real-time location of workers and equipment, and failing to meet the centimeter-level safety control requirements. In terms of safety monitoring, it can only identify obvious violations such as not wearing a safety helmet, lacking the ability to predict hidden risks such as abnormal work posture and abnormal equipment temperature, and cannot prevent safety accidents caused by misoperation in advance. In terms of data processing and transmission, it relies on centralized cloud processing, resulting in serious delays during large-scale operations. Moreover, traditional communication methods cannot guarantee the differentiated transmission needs of control signals and video streams, leading to low efficiency in collaborative operations. In terms of scenario adaptability, existing systems are mostly designed for single work scenarios, lacking a unified adaptation mechanism, and are unable to flexibly cope with the operational safety requirements of different environments such as power transmission towers and underground substations. In terms of risk intervention, the intervention methods are singular, and a graded response mechanism has not been established. It is impossible to take precise protective measures according to the risk level, either over-intervention affects operational efficiency or untimely intervention leads to safety accidents. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent safety protection system and method for power operations to solve the problems existing in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent safety protection system for power operations, comprising a multi-source fusion positioning and 3D modeling module, a safety status perception and risk prediction module, a hierarchical intervention control module, an edge computing and digital twin module, a 5G slice communication transmission module, a multi-scenario adaptation module, and a system data management and optimization module; The multi-source fusion positioning and 3D modeling module is used to acquire positioning data of operators and equipment and information about the working environment, construct a dynamic 3D model and correct positioning deviations. The safety status perception and risk prediction module is used to perceive the safety status of the work site, identify explicit violations, and predict the risk of implicit misoperation. The graded intervention control module is used to implement differentiated safety intervention measures according to the risk level; The edge computing and digital twin module is used to perform local data processing and simulate work processes through a digital twin model to identify potential security risks. The 5G slicing communication transmission module is used to ensure differentiated transmission of control signals and video streams through 5G slicing technology; The multi-scenario adaptation module is used to adapt to different types of power operation scenarios; The system data management and optimization module is used to store and analyze system operation data and optimize system performance.

[0005] Preferably, the multi-source fusion positioning and 3D modeling module includes a UWB positioning unit, a BeiDou positioning unit, a visual reconstruction unit, a dynamic 3D modeling unit, and a positioning deviation correction unit; UWB positioning unit: Through UWB base stations deployed at the work site and UWB tags carried by workers and equipment, it can achieve high-precision location acquisition over short distances and capture real-time displacement information of the work target. Beidou Positioning Unit: Receives Beidou satellite navigation signals and obtains global location data of the work area, providing basic support for positioning in open outdoor scenarios; Visual reconstruction unit: It collects environmental images and dynamic images of the target work by deploying high-definition cameras at the work site, extracts feature point information from the images, and provides data source for 3D modeling; Dynamic 3D Modeling Unit: The unit fuses image feature points collected by the visual reconstruction unit with multi-source positioning data to construct a dynamic 3D model of the work environment, personnel and equipment, reflecting the spatial changes of the work scene in real time. Positioning Deviation Correction Unit: Based on the spatial constraints of the dynamic 3D model, this unit corrects the raw UWB and BeiDou positioning data using a deviation correction algorithm. This eliminates positioning errors caused by obstructions and signal interference, outputting accurate positioning results. The deviation correction algorithm formula is as follows:

[0006] in, The final positioning coordinates after correction are in three-dimensional coordinate form, including the X, Y, and Z dimensions; The weighting coefficients for UWB positioning data are dynamically adjusted based on the UWB signal strength at the work site. The original positioning coordinates collected by the UWB positioning unit; The weighting coefficients for BeiDou positioning data are dynamically adjusted based on the quality of satellite signal reception. The original positioning coordinates collected by the BeiDou positioning unit; The weighting coefficients for visual reconstruction and localization data are dynamically adjusted based on the image feature matching degree. The location coordinates of the visual reconstruction unit are calculated based on image features; The deviation compensation provided for the dynamic 3D modeling unit is calculated from the spatial deviation between the positioning point in the 3D model and the feature point in the actual scene.

[0007] Preferably, the security status perception and risk prediction module includes an infrared thermal imaging perception unit, a millimeter-wave radar perception unit, a YOLOv8 optimized algorithm processing unit, an explicit violation identification unit, and a latent risk prediction unit; Infrared thermal imaging sensing unit: Collects temperature data of operating equipment and power transmission lines through an infrared thermal imager to detect abnormal heating phenomena of equipment; Millimeter-wave radar sensing unit: Utilizing the penetrating and anti-interference capabilities of millimeter-wave radar, it collects data on the operator's posture and movement trajectory. The YOLOv8 optimization algorithm processing unit: performs collaborative processing on infrared thermal imaging data, millimeter-wave radar data, and visual image data to optimize target detection accuracy and response speed, and improve feature recognition capabilities in complex environments; the specific calculation formula is as follows:

[0008] in, The final target detection confidence score; Indicates the number of data types, in this scenario =3, corresponding to infrared thermal imaging data, millimeter-wave radar data, and visual image data; For the first Weights of different data types, 0 ≤ ≤1, and The weights reflect the importance of this type of data in target detection; For the first The target detection confidence scores obtained after processing various data types using the YOLOv8 algorithm.

[0009] Explicit Violation Identification Unit: Based on the detection results of the optimized YOLOv8 algorithm, it identifies explicit security violations. The identification of explicit security violations is achieved through the following process based on the optimized YOLOv8 algorithm: Data preparation: Collect a large amount of image and video data from power operation sites, covering different operation scenarios, lighting conditions and personnel actions. Use annotation tools to accurately annotate safety-related elements in the data, such as wearing safety helmets, wearing insulating gloves, and using protective measures for working at heights, and build a training dataset. Algorithm optimization: The YOLOv8 algorithm was improved to suit the characteristics of power operation scenarios. For example, the network structure was adjusted to adapt to the features of safety violations. Transfer learning was used to pre-train model parameters on publicly available datasets, followed by fine-tuning using labeled power operation data, thereby improving the accuracy and speed of detecting safety violations. Real-time detection: The optimized YOLOv8 model is deployed to the intelligent safety protection system for power operations. The system analyzes the images and video streams of the work site collected in real time by the camera frame by frame. The algorithm identifies the workers and equipment through feature extraction, target classification and localization, and determines whether there are obvious safety violations such as not wearing a safety helmet or operating equipment in violation of regulations. Result determination: Set thresholds and rules for determining violations. If the detected target features match the violation pattern and the confidence level exceeds the threshold, the system determines it as an explicit security violation and triggers an alarm mechanism. At the same time, the violation information is recorded in the system data management module. Hidden Risk Prediction Unit: By combining operational posture data, equipment temperature data, and operational procedure specifications, a risk prediction model is established to identify hidden risks that may lead to misoperation; By combining explicit violations and implicit risks, a risk assessment index formula is constructed:

[0010] in, This is a safety risk assessment index, with a value range of [0, 10]. The larger the value, the higher the risk. This is the explicit violation weighting coefficient, set according to the severity of violations in the power safety regulations; Points are awarded for explicit violations, with different point values ​​for violations such as not wearing a safety helmet or not wearing insulated clothing. No points are awarded for no violations. This is the weighting coefficient for abnormal equipment temperature, set according to the equipment type and rated temperature. The score for abnormal equipment temperature is calculated from the difference between the actual equipment temperature and the rated temperature. A score of 0 is given when the temperature is normal. The abnormal work posture weighting coefficient is set according to the work scenario and safety operation specifications; The score for abnormal working posture is calculated based on the degree of deviation between the working posture and the standard posture. A score of 0 is given when the posture is normal.

[0011] Preferably, the graded intervention control module includes a risk level determination unit, an audible and visual early warning unit, an emergency stop control unit, and an intervention feedback unit; Risk level assessment unit: Based on the severity of explicit violations and the probability of occurrence of implicit risks, safety risks are divided into three levels: low, medium, and high. Audible and visual early warning unit: When the risk level is low or medium, the audible and visual alarm will be activated, emitting alarm sounds and warning lights at different frequencies to remind workers to correct violations or adjust their working posture in a timely manner. Emergency stop control unit: When the risk level is high, it sends an emergency stop command to the working equipment and power line control terminal to cut off the working power supply or stop the equipment operation to avoid safety accidents. Intervention Feedback Unit: Records the implementation status of intervention measures and transmits feedback information to the system data management and optimization module to provide data support for subsequent risk prediction model optimization.

[0012] Preferably, the edge computing and digital twin module includes an edge node processing unit, a digital twin modeling unit, a workflow simulation unit, and a hazard identification unit; Edge node processing unit: Based on the power IoT operating system architecture, edge computing nodes are deployed at the work site to process data collected by multiple sensors locally, reduce data transmission latency, and improve real-time response speed; Digital twin modeling unit: Based on the geographical information, equipment parameters, and work process specifications of the work site, construct a digital twin model that corresponds to the physical work scene in a 1:1 ratio, and realize the virtual mapping of the work scene; Workflow simulation unit: Simulates the entire work process in a digital twin model, simulates personnel movement and equipment operation status under different work steps, and predicts possible spatial conflicts, operational conflicts and other problems. Hazard identification unit: By comparing the simulated work process with safety regulations, it identifies safety hazards in the process design, outputs the location and type of hazards and rectification suggestions, and provides support for optimizing the work plan.

[0013] Preferably, the 5G slice communication transmission module includes a 5G slice configuration unit, a control signal transmission unit, a video stream transmission unit, and a transmission quality monitoring unit; The 5G slicing configuration unit is used to divide the data into dedicated slices for control signals and dedicated slices for video streams; based on data transmission priority and real-time data volume, a 5G slice bandwidth allocation formula is constructed:

[0014] in, For the first Bandwidth allocation for slices corresponding to data types =1 indicates a control signal. When the value is 2, it is a video stream; For the first Priority coefficients for data types, with control signals having a higher priority coefficient than video streams; The base bandwidth is set based on the total available bandwidth of the 5G network at the work site; For the first The bandwidth adjustment factor for this type of data is set according to the real-time requirements of data transmission; For the first The real-time transmission volume of this type of data is obtained from the real-time statistics of the edge computing nodes; The control signal transmission unit is used to transmit key control data; The video stream transmission unit is used to transmit real-time video data from the work site. The transmission quality monitoring unit is used to monitor transmission quality and dynamically adjust network resource allocation.

[0015] Preferably, the multi-scene adaptation module includes a scene recognition unit, a parameter configuration unit, and a device adaptation unit; Scene recognition unit: By matching positioning data, environmental image features and a preset scene feature library, it automatically identifies the type of the current work scene, such as high-altitude work on power transmission towers, underground power distribution room work, etc. Parameter configuration unit: Automatically retrieves the system configuration information corresponding to the identified scene type; Equipment adapter unit: Supports communication adaptation with different types of power operation equipment, sensors, and control terminals, and realizes data interaction through standardized interfaces to improve system compatibility.

[0016] Preferably, the system data management and optimization module includes a data storage unit, a data analysis unit, a model optimization unit, and a historical data retrieval unit; Data storage unit: Classifies and stores location data, safety status perception data, risk assessment data, intervention execution data, and work process data to build a complete system operation database; Data Analysis Unit: Performs statistical analysis on stored data to uncover patterns in risk occurrence, risk characteristics in different scenarios, and weaknesses in system operation; Model optimization unit: Based on data analysis results, adjust the parameters of the positioning deviation correction algorithm, the weight coefficients of the risk prediction model, and the 5G slicing resource allocation strategy to continuously improve system performance; Historical data retrieval unit: Provides historical data support for digital twin operation simulation and parameter configuration of new operation scenarios, and assists in the formulation of operation plans.

[0017] A protection method for an intelligent safety protection system for power operations includes the following steps: Step 1: System initialization and scene adaptation, each module performs self-test initialization, and the multi-scene adaptation module completes scene recognition, parameter configuration and device adaptation; Step 2: Multi-source fusion localization and 3D modeling, collecting multi-source localization data and image features, constructing a dynamic 3D model and correcting localization deviations; Step 3: Safety status perception and risk prediction, collect data from multiple sensors, identify explicit violations and predict implicit risks, and calculate the risk level; Step 4: Edge computing and digital twin risk assessment, local data processing, construction of digital twin model and simulation of work process to identify potential security risks; Step 5: Ensure 5G slicing communication transmission, transmit various types of data through 5G slicing, and monitor and adjust transmission quality; Step 6: Implement tiered intervention and control measures, and execute audible and visual warnings or emergency shutdown interventions based on the risk level, and record intervention feedback; Step 7: Data management and system optimization, storing and analyzing system operation data, optimizing system performance and archiving data; Step 8: The operation ends and the system is reset. All modules stop working and are reset to standby state.

[0018] The beneficial effects of this invention are: This invention employs a multi-source fusion positioning technology combining UWB, BeiDou, and visual reconstruction, along with dynamic 3D modeling for positioning deviation correction. This effectively overcomes the limitations of single technologies in complex environments, achieving high-precision positioning and providing accurate location support for safety protection. By integrating infrared thermal imaging, millimeter-wave radar, and the YOLOv8 optimization algorithm, it can identify explicit safety violations and predict hidden operational risks based on data such as equipment temperature and operating posture, achieving comprehensive safety status awareness. A tiered intervention mechanism is established, implementing differentiated audible and visual warnings or emergency shutdown measures based on risk levels. This ensures operational safety while avoiding excessive intervention that could impact efficiency, achieving a balance between safety and efficiency. Local data processing is achieved through edge computing nodes, reducing transmission latency. Combined with 5G slicing technology, differentiated transmission of control signals and video streams is ensured, resolving data latency and coordination challenges during large-scale operations and improving the system's real-time response capabilities. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1: This invention provides an intelligent safety protection system for power operations, including a multi-source fusion positioning and 3D modeling module, a safety status perception and risk prediction module, a hierarchical intervention control module, an edge computing and digital twin module, a 5G slice communication transmission module, a multi-scenario adaptation module, and a system data management and optimization module; The multi-source fusion positioning and 3D modeling module is the core of the system's precise positioning. It is used to acquire real-time location data of operators and equipment, as well as information about the working environment. Through dynamic 3D modeling, it corrects positioning deviations and ensures that positioning accuracy is adapted to complex working scenarios.

[0022] This module specifically includes a UWB positioning unit, a BeiDou positioning unit, a visual reconstruction unit, a dynamic 3D modeling unit, and a positioning deviation correction unit. UWB positioning unit: Through UWB base stations deployed at the work site and UWB tags carried by workers and equipment, it can achieve high-precision location acquisition over short distances and capture real-time displacement information of the work target. Beidou Positioning Unit: Receives Beidou satellite navigation signals and obtains global location data of the work area, providing basic support for positioning in open outdoor scenarios; Visual reconstruction unit: It collects environmental images and dynamic images of the target work by deploying high-definition cameras at the work site, extracts feature point information from the images, and provides data source for 3D modeling; Dynamic 3D Modeling Unit: The unit fuses image feature points collected by the visual reconstruction unit with multi-source positioning data to construct a dynamic 3D model of the work environment, personnel and equipment, reflecting the spatial changes of the work scene in real time. Positioning Deviation Correction Unit: Based on the spatial constraints of the dynamic 3D model, this unit corrects the raw UWB and BeiDou positioning data using a deviation correction algorithm. This eliminates positioning errors caused by factors such as obstruction and signal interference, outputting accurate positioning results. The deviation correction algorithm formula is as follows:

[0023] in, The final positioning coordinates after correction are in three-dimensional coordinate form, including the X, Y, and Z dimensions; The weighting coefficients for UWB positioning data are dynamically adjusted based on the UWB signal strength at the work site. The original positioning coordinates collected by the UWB positioning unit; The weighting coefficients for BeiDou positioning data are dynamically adjusted based on the quality of satellite signal reception. The original positioning coordinates collected by the BeiDou positioning unit; The weighting coefficients for visual reconstruction and localization data are dynamically adjusted based on the image feature matching degree. The location coordinates of the visual reconstruction unit are calculated based on image features; The deviation compensation provided for the dynamic 3D modeling unit is calculated from the spatial deviation between the positioning point in the 3D model and the feature point in the actual scene.

[0024] The safety status perception and risk prediction module is used to comprehensively perceive the safety status of the work site, identify both explicit violations and predict implicit risks of misoperation, and provide a basis for graded intervention.

[0025] This module specifically includes an infrared thermal imaging sensing unit, a millimeter-wave radar sensing unit, a YOLOv8 optimized algorithm processing unit, an explicit violation identification unit, and a latent risk prediction unit. Infrared thermal imaging sensing unit: Collects temperature data of operating equipment and power transmission lines through an infrared thermal imager to detect abnormal heating phenomena of equipment; Millimeter-wave radar sensing unit: Utilizing the penetrating and anti-interference capabilities of millimeter-wave radar, it collects data on the worker's posture and movement trajectory, unaffected by environmental factors such as light and smoke; The YOLOv8 optimization algorithm processing unit: performs collaborative processing on infrared thermal imaging data, millimeter-wave radar data, and visual image data to optimize target detection accuracy and response speed, and improve feature recognition capabilities in complex environments; the specific calculation formula is as follows:

[0026] in, The final target detection confidence score; Indicates the number of data types, in this scenario =3, corresponding to infrared thermal imaging data, millimeter-wave radar data, and visual image data; For the first Weights of different data types, 0 ≤ ≤1, and The weights reflect the importance of this type of data in target detection; For the first The target detection confidence scores obtained after processing various data types using the YOLOv8 algorithm; Obvious violation identification unit: Based on the detection results of the YOLOv8 optimized algorithm, it identifies obvious safety violations such as not wearing a safety helmet or not wearing insulated clothing; Based on the optimized YOLOv8 algorithm, the identification of explicit security violations is achieved through the following process: Data preparation: Collect a large amount of image and video data from power operation sites, covering different operation scenarios, lighting conditions and personnel actions. Use annotation tools to accurately annotate safety-related elements in the data, such as wearing safety helmets, wearing insulating gloves, and using protective measures for working at heights, and build a training dataset. Algorithm optimization: The YOLOv8 algorithm was improved to suit the characteristics of power operation scenarios. For example, the network structure was adjusted to adapt to the features of safety violations. Transfer learning was used to pre-train model parameters on publicly available datasets, followed by fine-tuning using labeled power operation data, thereby improving the accuracy and speed of detecting safety violations. Real-time detection: The optimized YOLOv8 model is deployed to the intelligent safety protection system for power operations. The system analyzes the images and video streams of the work site collected in real time by the camera frame by frame. The algorithm identifies the workers and equipment through feature extraction, target classification and localization, and determines whether there are obvious safety violations such as not wearing a safety helmet or operating equipment in violation of regulations. Result determination: Set thresholds and rules for determining violations. If the detected target features match the violation pattern and the confidence level exceeds the threshold, the system determines it as an explicit security violation and triggers an alarm mechanism. At the same time, the violation information is recorded in the system data management module. Hidden Risk Prediction Unit: Combining work posture data, equipment temperature data, and work procedure specifications, a risk prediction model is established to identify hidden risks that may lead to misoperation, such as excessive bending, unauthorized approach to live parts, and abnormal rise in equipment temperature. The risk prediction model is established through the following steps, combining work posture data, equipment temperature data, and work procedure specifications: Data Acquisition and Preprocessing: Wearable devices and sensor networks are used to collect real-time posture data of workers (such as range of motion, angle changes, and movement trajectories) and equipment temperature data. The raw data is cleaned, noise reduced, and normalized. At the same time, the work process specifications are transformed into structured data, and the standard operating parameters and safety thresholds of each link are clearly defined.

[0027] Feature extraction and correlation analysis: Key features are extracted from processed data using machine learning algorithms. For example, abnormal movement patterns are extracted from operational posture data, and temperature change trends are analyzed from equipment temperature data. Combined with operational process specifications, correlations between data are established to clarify the impact weight of different combinations of data features on safety risks in each operational stage.

[0028] Model Building and Training: A risk prediction model is built using deep learning (such as LSTM, Transformer, etc.) or traditional machine learning algorithms (such as random forest, support vector machine). Historical job data and corresponding risk events are used as training samples to train the model and optimize its parameters, enabling the model to accurately identify the mapping relationship between data features and risks.

[0029] Model Validation and Optimization: The trained model is validated using an independent test dataset to evaluate metrics such as accuracy and recall. Based on feedback from real-world application scenarios, the model structure and parameters are continuously adjusted, and the risk prediction logic is optimized to ensure the model can accurately identify potential hidden risks that could lead to erroneous operations.

[0030] By combining explicit violations and implicit risks, a risk assessment index formula is constructed:

[0031] in, This is a safety risk assessment index, with a value range of [0, 10]. The larger the value, the higher the risk. This is the explicit violation weighting coefficient, set according to the severity of violations in the power safety regulations; Points are awarded for explicit violations, with different point values ​​for violations such as not wearing a safety helmet or not wearing insulated clothing. No points are awarded for no violations. This is the weighting coefficient for abnormal equipment temperature, set according to the equipment type and rated temperature. The score for abnormal equipment temperature is calculated from the difference between the actual equipment temperature and the rated temperature. A score of 0 is given when the temperature is normal. The abnormal work posture weighting coefficient is set according to the work scenario and safety operation specifications; The score for abnormal working posture is calculated based on the degree of deviation between the working posture and the standard posture. A score of 0 is given when the posture is normal.

[0032] The graded intervention and control module is used to execute differentiated safety intervention measures based on the risk level output by the safety status perception and risk prediction module, so as to achieve precise protection.

[0033] This module specifically includes a risk level determination unit, an audible and visual early warning unit, an emergency shutdown control unit, and an intervention feedback unit. Risk Level Assessment Unit: Based on the severity of explicit violations and the probability of occurrence of implicit risks, safety risks are divided into three levels: low, medium, and high. Low risk: Overt violations are minor operational errors, such as not wearing some non-critical protective equipment correctly, which do not cause direct safety hazards; latent risks have a probability of less than 10%, such as equipment aging but no risk of failure in the short term, which can be eliminated by simple rectification.

[0034] Medium risk: Overt violations involve breaches of important safety regulations, such as failure to follow procedures for powering off equipment, which poses a potential safety threat; latent risks have a probability of occurrence between 10% and 50%, such as a decrease in the local insulation performance of the equipment, which may lead to malfunctions, and require rectification within a specified period.

[0035] High risk: Overt violations are serious breaches of safety regulations. Entering a high-risk work area without taking any protective measures directly threatens personnel safety. The probability of hidden risks exceeds 50%. For example, if critical components of equipment are seriously damaged, it may cause a major accident at any time. Work must be stopped immediately and a thorough inspection and repair must be carried out. Audible and visual early warning unit: When the risk level is low or medium, the audible and visual alarm will be activated, emitting alarm sounds and warning lights at different frequencies to remind workers to correct violations or adjust their working posture in a timely manner. Emergency stop control unit: When the risk level is high, it sends an emergency stop command to the working equipment and power line control terminal to cut off the working power supply or stop the equipment operation to avoid safety accidents. Intervention Feedback Unit: Records the implementation status of intervention measures and transmits feedback information to the system data management and optimization module to provide data support for subsequent risk prediction model optimization.

[0036] Edge computing and digital twin modules are used to improve data processing efficiency, identify potential security risks in advance, and ensure the rationality and security of work processes.

[0037] This module specifically includes an edge node processing unit, a digital twin modeling unit, a workflow simulation unit, and a hazard identification unit: Edge Node Processing Unit: Based on the power IoT operating system architecture, edge computing nodes are deployed at the work site to perform the following local processing on data collected by multiple sensors: Data preprocessing: Noise in sensor data is removed by filtering algorithms, and data format is unified by normalization methods to improve data quality; Feature extraction: Deep learning algorithms are used to extract key features from the raw data, such as equipment operating parameters and worker postures; Real-time analysis: Utilizing the intelligent algorithms built into edge computing nodes, the processed data is analyzed in real time to quickly identify abnormal states, such as equipment overheating or personnel violations. Decision execution: Based on the analysis results, corresponding security policies are triggered locally, such as issuing alarms and controlling equipment shutdown. At the same time, key information is synchronized to the cloud for in-depth analysis and storage, reducing data transmission latency and improving real-time response speed. Digital twin modeling unit: Based on the geographical information, equipment parameters, and work process specifications of the work site, construct a digital twin model that corresponds to the physical work scene in a 1:1 ratio, and realize the virtual mapping of the work scene; Workflow simulation unit: Simulates the entire work process in a digital twin model, simulates personnel movement and equipment operation status under different work steps, and predicts possible spatial conflicts, operational conflicts and other problems. Hazard identification unit: By comparing the simulated work process with safety regulations, it identifies safety hazards in the process design, outputs the location and type of hazards and rectification suggestions, and provides support for optimizing the work plan.

[0038] The 5G slicing communication transmission module is used to ensure the stable transmission of various types of data at the work site, meet the differentiated transmission needs of control signals and video streams, and solve the coordination problem during large-scale operations.

[0039] This module specifically includes a 5G slicing configuration unit, a control signal transmission unit, a video stream transmission unit, and a transmission quality monitoring unit: 5G Slicing Configuration Unit: Based on the data transmission requirements of power operations, the 5G network is divided into dedicated slices for control signals and dedicated slices for video streams, with independent network resources allocated to each, and a 5G slice bandwidth allocation formula is constructed:

[0040] in, For the first Bandwidth allocation for slices corresponding to data types =1 indicates a control signal. When the value is 2, it is a video stream; For the first Priority coefficients for data types, with control signals having a higher priority coefficient than video streams; The base bandwidth is set based on the total available bandwidth of the 5G network at the work site; For the first The bandwidth adjustment factor for this type of data is set according to the real-time requirements of data transmission; For the first The real-time transmission volume of this type of data is obtained from the real-time statistics of the edge computing nodes; Control signal transmission unit: Transmits critical data such as emergency stop commands and early warning signals through dedicated control signal slices, ensuring low latency and high reliability of transmission; Video Stream Transmission Unit: Transmits real-time video data from the work site through dedicated video stream slices, ensuring smooth video transmission and meeting remote monitoring requirements; Transmission quality monitoring unit: Real-time monitoring of transmission bandwidth, latency, packet loss rate and other indicators of the two slices. When the transmission quality does not meet the standards, it dynamically adjusts the allocation of network resources to ensure stable data transmission.

[0041] The multi-scenario adaptation module enables the system to flexibly adapt to different power operation scenarios, meeting diverse operational needs without separate configuration.

[0042] This module specifically includes a scene recognition unit, a parameter configuration unit, and a device adaptation unit: Scene recognition unit: By matching positioning data, environmental image features and a preset scene feature library, it automatically identifies the type of the current work scene, such as high-altitude work on power transmission towers, underground power distribution room work, etc. Parameter configuration unit: Based on the identified scene type, it automatically calls the corresponding positioning parameters, sensor acquisition parameters, risk prediction thresholds and other configuration information to optimize the system performance in specific scenarios; Equipment adapter unit: Supports communication adaptation with different types of power operation equipment, sensors, and control terminals, and realizes data interaction through standardized interfaces to improve system compatibility.

[0043] The system data management and optimization module is used to store and analyze various types of data during system operation, continuously optimize system performance, and improve the accuracy of security protection.

[0044] This module specifically includes a data storage unit, a data analysis unit, a model optimization unit, and a historical data retrieval unit: Data storage unit: Classifies and stores location data, safety status perception data, risk assessment data, intervention execution data, work process data, etc., to build a complete system operation database; Data Analysis Unit: Performs statistical analysis on stored data to uncover patterns in risk occurrence, risk characteristics in different scenarios, and weaknesses in system operation; Model optimization unit: Based on data analysis results, adjust the parameters of the positioning deviation correction algorithm, the weight coefficients of the risk prediction model, and the 5G slicing resource allocation strategy to continuously improve system performance; Historical data retrieval unit: Provides historical data support for digital twin operation simulation and parameter configuration of new operation scenarios, and assists in the formulation of operation plans.

[0045] Example 2: A protection method for an intelligent safety protection system for power operations, comprising the following steps: Step 1: System Initialization and Scene Adaptation like Figure 1 As shown, the intelligent safety protection system for power operations is activated, and each module completes self-checks and initialization. The scene recognition unit of the multi-scenario adaptation module collects the positioning data and environmental image features of the work site, matches them with the preset scene feature library, and automatically identifies the current work scenario type (such as high-altitude work on transmission towers, underground substation work, etc.). Based on the recognition results, the parameter configuration unit calls up the corresponding positioning parameters, sensor acquisition parameters, risk prediction thresholds, and other configuration information. The equipment adaptation unit completes communication adaptation with the work equipment, sensors, and control terminal, and the system enters standby mode.

[0046] Step 2: Multi-source fusion localization and 3D modeling The multi-source fusion positioning and 3D modeling module begins operation: the UWB positioning unit and the BeiDou positioning unit simultaneously collect raw positioning data of personnel and equipment; the visual reconstruction unit collects images of the work environment and dynamic images of the work target through high-definition cameras, and extracts image feature points; the dynamic 3D modeling unit fuses positioning data and image feature points to construct a dynamic 3D model of the work scene; the positioning deviation correction unit, based on the spatial constraints of the 3D model, corrects the raw positioning data through the positioning deviation correction formula, outputs positioning results with centimeter-level accuracy, and updates them to relevant modules of the system in real time.

[0047] Step 3: Safety Status Perception and Risk Prediction The safety status perception and risk prediction module initiates data acquisition and analysis: the infrared thermal imaging perception unit collects temperature data of the operating equipment and power transmission lines; the millimeter-wave radar perception unit collects data on the working posture and movement trajectory of the operators; the YOLOv8 optimization algorithm processing unit performs collaborative processing on the data from various sensors to improve target detection accuracy; the explicit violation identification unit identifies explicit violations based on the processing results; the implicit risk prediction unit combines equipment temperature data, working posture data, and operating procedure specifications to predict implicit misoperation risks; and the risk assessment index is calculated using the safety risk assessment index formula, and the risk level determination unit determines the risk level, specifically low risk, medium risk, and high risk.

[0048] Step 4: Risk Assessment of Edge Computing and Digital Twins Edge computing and digital twin modules work in parallel: the edge node processing unit processes local positioning and status perception data collected by multi-source sensors to reduce data transmission latency; the digital twin modeling unit constructs a digital twin model based on the geographic information, equipment parameters, and initial configuration of the work site; the work process simulation unit simulates the entire work process in the digital twin model, simulating personnel movement and equipment operating status; the hazard identification unit compares the simulated process with safety specifications, identifies safety hazards in the process design, outputs rectification suggestions, and promptly reports any major hazards to the work command terminal, resuming work only after the hazards have been rectified.

[0049] Step 5: 5G Slicing Communication Transmission Guarantee The 5G slicing configuration unit of the 5G slicing communication transmission module divides the 5G network into dedicated control signal slices and dedicated video stream slices, allocating independent network resources. During operations, the control signal transmission unit transmits key data such as positioning results, risk levels, and intervention commands through the dedicated control signal slice, while the video stream transmission unit transmits real-time video data from the operation site through the dedicated video stream slice. The transmission quality monitoring unit monitors the transmission bandwidth, latency, packet loss rate, and other indicators of both slices in real time. When transmission quality fails to meet standards, the bandwidth allocation is dynamically adjusted using the 5G slice bandwidth allocation formula to ensure stable data transmission.

[0050] Step 6: Implementation of Tiered Intervention and Control The tiered intervention control module executes differentiated intervention measures based on the risk level output in step 3: If the risk level is low, the audible and visual warning unit will activate a low-frequency audible and visual alarm to remind workers to pay attention to standardized operations; If the risk level is medium, the audible and visual warning unit will activate a high-frequency audible and visual alarm and send a warning message to the operation command terminal. Commanders can then remotely guide workers to correct any violations. If the risk level is high, the emergency stop control unit will immediately send an emergency stop command to the working equipment and power line control terminal to cut off the working power supply or stop the equipment operation to avoid safety accidents. The intervention feedback unit records the implementation status of intervention measures and transmits the feedback information to the system data management and optimization module.

[0051] Step 7: Data Management and System Optimization The system data management and optimization module continuously stores location data, status awareness data, risk assessment data, and intervention execution data during the operation process; the data analysis unit performs statistical analysis on the stored data to uncover patterns of risk occurrence and weaknesses in system operation; the model optimization unit adjusts the parameters of the positioning deviation correction algorithm, the weight coefficients of the risk prediction model, and the 5G slice resource allocation strategy based on the analysis results to optimize system performance; and the historical data retrieval unit archives and stores valid data to provide data support for scenario adaptation, process simulation, and hidden danger investigation in subsequent operations.

[0052] Step 8: Operation completed and system reset Once the power operation is completed, the operator sends an operation completion command. The system then stops data acquisition and risk monitoring, the hierarchical intervention control module disables the early warning and shutdown control functions, the edge computing and digital twin module stops simulation and troubleshooting, the system data management and optimization module completes final data archiving and model optimization, and all modules reset to standby mode, awaiting the next operation command.

[0053] System Deployment and Initialization: At the power transmission tower operation site, the intelligent safety protection system for power operations described in this invention is deployed. The scene recognition unit of the multi-scene adaptation module collects the geographical features of the power transmission tower and images of the high-altitude operation environment, matches them with a preset power transmission tower operation scene feature library, and automatically identifies the scene type. The parameter configuration unit calls the positioning parameters and sensor acquisition parameters corresponding to the high-altitude operation, and the equipment adaptation unit completes communication adaptation with the tower operation equipment, sensors, and power transmission line control terminal. The system completes initialization and enters standby mode.

[0054] Location and modeling process: After the multi-source fusion positioning and 3D modeling module is activated, operators wear UWB tags, and the operating equipment is equipped with UWB tags and BeiDou positioning modules to simultaneously collect location data. The visual reconstruction unit uses high-definition cameras deployed at the bottom and middle of the tower to collect dynamic images of the tower structure, operator movements, and the surrounding environment, extracting feature points such as tower nodes and operator outlines. The dynamic 3D modeling unit integrates UWB and BeiDou positioning data and image feature points to construct a dynamic 3D model of the transmission tower and its surrounding environment, presenting the real-time location of operators and the distribution of equipment. The positioning deviation correction unit, based on the spatial constraints of the tower structure in the 3D model, eliminates positioning errors caused by high-altitude signal obstruction through a positioning deviation correction formula, outputting accurate positioning results to ensure that the distance monitoring accuracy between operators and live conductors meets safety requirements.

[0055] Safety perception and risk prediction: In the safety status perception and risk prediction module, the infrared thermal imaging sensing unit continuously collects temperature data by targeting key parts such as transmission line joints and insulators; the millimeter-wave radar sensing unit is installed in the middle of the tower to scan the workers' postures, such as climbing movements and body tilt angles during operation. The YOLOv8 optimization algorithm processing unit performs collaborative analysis on the temperature data, posture data, and image data collected by the camera. The explicit violation identification unit monitors in real time whether workers are wearing safety helmets and insulated clothing; if not, it is identified as an explicit violation. The implicit risk prediction unit combines work posture data (such as excessive bending may lead to instability) and equipment temperature data (such as abnormally high joint temperatures may cause short circuits) to predict the risk of misoperation. The risk index is calculated using the safety risk assessment index formula. If a worker is not wearing a safety helmet and is close to a live conductor, the risk index reaches a high-level threshold, and the risk level determination unit outputs a high-risk level.

[0056] Edge computing and digital twin applications: Edge computing nodes are deployed in control boxes at the work site to process positioning, temperature, and attitude data locally, avoiding delays in data transmission to the remote cloud. The digital twin modeling unit constructs a digital twin model based on the design parameters and work plan of the transmission tower. The work process simulation unit simulates the entire process of personnel climbing, equipment installation, and line maintenance within the model, simulating the positional relationship between personnel and equipment at different work steps. The hazard identification unit compares the simulated process with safety regulations for high-altitude power operations, identifying the hazard of "potential spatial conflict between workers and the lines below when working at a certain height on the tower," and outputs rectification suggestions to adjust the work sequence. Work supervisors optimize the work plan based on these suggestions to avoid safety risks during actual operations.

[0057] Communication transmission and hierarchical intervention: The 5G slicing communication transmission module divides the 5G network into control signal slices and video stream slices. Control signal slices transmit critical data such as positioning results, risk levels, and emergency stop commands, while video stream slices transmit real-time video of personnel working at heights. The transmission quality monitoring unit monitors the transmission status of each slice in real time. When fluctuations in the high-altitude signal cause insufficient bandwidth for the video stream transmission, the bandwidth is dynamically adjusted using the 5G slice bandwidth allocation formula to ensure smooth video stream transmission and facilitate real-time monitoring by ground command personnel. When the risk level is determined to be high, the emergency stop control unit of the tiered intervention control module sends an emergency stop command to the power transmission line control terminal, cutting off the power supply to the work area. Simultaneously, the audible and visual warning unit activates a high-frequency alarm to remind personnel to suspend work. The intervention feedback unit records the command execution status and feeds it back to the system data management and optimization module.

[0058] Data Management and System Optimization: The system data management and optimization module stores location data, temperature data, risk assessment data, and intervention execution data for this operation. The data analysis unit found that the UWB positioning deviation was slightly higher in a specific height area of ​​the tower than in other areas. The model optimization unit adjusted the weighting coefficient of the visual reconstruction data in the positioning deviation correction formula for this area, improving positioning accuracy. The historical data retrieval unit archives data such as the hazard investigation results and the effectiveness of intervention measures for this operation, providing a reference for scenario adaptation and process optimization for subsequent similar transmission tower operations.

[0059] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0060] Example 3: A terminal, comprising a processor and a storage medium; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method.

[0061] Example 4: A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the method.

[0062] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0063] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0064] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0065] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. An intelligent safety protection system for power operations, characterized in that: It includes a multi-source fusion positioning and 3D modeling module, a security status perception and risk prediction module, a hierarchical intervention and control module, an edge computing and digital twin module, a 5G slicing communication transmission module, a multi-scenario adaptation module, and a system data management and optimization module; The multi-source fusion positioning and 3D modeling module is used to acquire positioning data of operators and equipment and information about the working environment, construct a dynamic 3D model and correct positioning deviations. The safety status perception and risk prediction module is used to perceive the safety status of the work site, identify explicit violations, and predict the risk of implicit misoperation. The graded intervention control module is used to implement differentiated safety intervention measures according to the risk level; The edge computing and digital twin module is used to perform local data processing and to simulate work processes using a digital twin model to identify potential security risks. The 5G slicing communication transmission module is used to ensure differentiated transmission of control signals and video streams through 5G slicing technology; The multi-scenario adaptation module is used to adapt to different types of power operation scenarios; The system data management and optimization module is used to store and analyze system operation data and optimize system performance.

2. The intelligent safety protection system for power operations according to claim 1, characterized in that: The multi-source fusion positioning and 3D modeling module includes a UWB positioning unit, a BeiDou positioning unit, a visual reconstruction unit, a dynamic 3D modeling unit, and a positioning deviation correction unit. UWB positioning unit: Through UWB base stations deployed at the work site and UWB tags carried by workers and equipment, it can achieve high-precision location acquisition over short distances and capture real-time displacement information of the work target. Beidou Positioning Unit: Receives Beidou satellite navigation signals and obtains global location data of the work area, providing basic support for positioning in open outdoor scenarios; Visual reconstruction unit: It collects environmental images and dynamic images of the target work by deploying high-definition cameras at the work site, extracts feature point information from the images, and provides data source for 3D modeling; Dynamic 3D Modeling Unit: The unit fuses image feature points collected by the visual reconstruction unit with multi-source positioning data to construct a dynamic 3D model of the work environment, personnel and equipment, reflecting the spatial changes of the work scene in real time. Positioning Deviation Correction Unit: Based on the spatial constraints of the dynamic 3D model, the unit corrects the raw data of UWB and BeiDou positioning using a deviation correction algorithm to output accurate positioning results. The deviation correction algorithm formula is as follows: multiply the raw positioning coordinates collected by the UWB positioning unit by their corresponding weight coefficients, add the product of the raw positioning coordinates collected by the BeiDou positioning unit and their corresponding weight coefficients, add the product of the positioning coordinates calculated by the visual reconstruction unit based on image features and their corresponding weight coefficients, and finally add the deviation compensation provided by the dynamic 3D modeling unit to obtain the corrected final positioning coordinates.

3. The intelligent safety protection system for power operations according to claim 1, characterized in that: The security status perception and risk prediction module includes an infrared thermal imaging perception unit, a millimeter-wave radar perception unit, a YOLOv8 optimized algorithm processing unit, an explicit violation identification unit, and a hidden risk prediction unit. Infrared thermal imaging sensing unit: Collects temperature data of operating equipment and power transmission lines through an infrared thermal imager to detect abnormal heating phenomena of equipment; Millimeter-wave radar sensing unit: Utilizing the penetrating and anti-interference capabilities of millimeter-wave radar, it collects data on the operator's posture and movement trajectory. The YOLOv8 optimization algorithm processing unit: performs collaborative processing on infrared thermal imaging data, millimeter-wave radar data, and visual image data to optimize target detection accuracy and response speed, and improve feature recognition capabilities in complex environments; the specific calculation formula is as follows: The final target detection confidence score is calculated using a weighted summation method. This involves multiplying the corresponding weight of each data type by the target detection confidence score obtained after processing the data type using the YOLOv8 algorithm, and then summing the product results for all data types to obtain the final target detection confidence score. There are three types of data, corresponding to infrared thermal imaging data, millimeter-wave radar data, and visual image data, respectively. The weight reflects the importance of this type of data in target detection, and the sum of the weights of all data types is equal to one. The value of each weight ranges from zero to one. Explicit violation identification unit: Based on the detection results of the YOLOv8 optimized algorithm, it identifies explicit security violations; Hidden Risk Prediction Unit: By combining operational posture data, equipment temperature data, and operational procedure specifications, a risk prediction model is established to identify hidden risks that may lead to misoperation; The specific calculation method of the risk assessment index is as follows: multiply the explicit violation weight coefficient by the explicit violation score to obtain the first product, multiply the equipment temperature abnormality weight coefficient by the equipment temperature abnormality score to obtain the second product, multiply the work posture abnormality weight coefficient by the work posture abnormality score to obtain the third product; finally, add the first product, the second product and the third product to obtain the safety risk assessment index.

4. The intelligent safety protection system for power operations according to claim 1, characterized in that: The graded intervention control module includes a risk level determination unit, an audible and visual early warning unit, an emergency stop control unit, and an intervention feedback unit. Risk level assessment unit: Based on the severity of explicit violations and the probability of occurrence of implicit risks, safety risks are divided into three levels: low, medium, and high. Audible and visual early warning unit: When the risk level is low or medium, the audible and visual alarm will be activated, emitting alarm sounds and warning lights at different frequencies to remind workers to correct violations or adjust their working posture in a timely manner. Emergency stop control unit: When the risk level is high, it sends an emergency stop command to the working equipment and power line control terminal to cut off the working power supply or stop the equipment operation to avoid safety accidents. Intervention Feedback Unit: Records the implementation status of intervention measures and transmits feedback information to the system data management and optimization module to provide data support for subsequent risk prediction model optimization.

5. The intelligent safety protection system for power operations according to claim 1, characterized in that: The edge computing and digital twin module includes an edge node processing unit, a digital twin modeling unit, a workflow simulation unit, and a hazard identification unit; Edge node processing unit: Based on the power IoT operating system architecture, edge computing nodes are deployed at the work site to process data collected by multiple sensors locally; Digital twin modeling unit: Based on the geographical information of the work site, equipment parameters, and work process specifications, construct a digital twin model corresponding to the physical work scenario; Workflow simulation unit: Simulates the entire work process in a digital twin model, simulating personnel movement and equipment operation status under different work steps; Hazard identification unit: By comparing the simulated work process with safety regulations, identify safety hazards in the process design and output the location, type, and rectification suggestions of the hazards.

6. The intelligent safety protection system for power operations according to claim 1, characterized in that: The 5G slice communication transmission module includes a 5G slice configuration unit, a control signal transmission unit, a video stream transmission unit, and a transmission quality monitoring unit. The 5G slicing configuration unit is used to divide the dedicated slices for control signals and the dedicated slices for video streams; The control signal transmission unit is used to transmit key control data; The video stream transmission unit is used to transmit real-time video data from the work site. The transmission quality monitoring unit is used to monitor transmission quality and dynamically adjust network resource allocation.

7. The intelligent safety protection system for power operations according to claim 1, characterized in that: The multi-scene adaptation module includes a scene recognition unit, a parameter configuration unit, and a device adaptation unit; Scene recognition unit: By matching positioning data, environmental image features, and a preset scene feature library, it automatically identifies the type of the current work scene; Parameter configuration unit: Automatically retrieves the system configuration information corresponding to the identified scene type; Equipment adapter unit: Supports communication adaptation with different types of power operation equipment, sensors, and control terminals.

8. The intelligent safety protection system for power operations according to claim 1, characterized in that: The system data management and optimization module includes a data storage unit, a data analysis unit, a model optimization unit, and a historical data retrieval unit; Data storage unit: Classifies and stores location data, safety status perception data, risk assessment data, intervention execution data, and work process data to build a complete system operation database; Data Analysis Unit: Performs statistical analysis on stored data to uncover patterns in risk occurrence, risk characteristics in different scenarios, and weaknesses in system operation; Model optimization unit: Based on data analysis results, adjust the parameters of the positioning deviation correction algorithm, the weight coefficients of the risk prediction model, and the 5G slicing resource allocation strategy to continuously improve system performance; Historical data retrieval unit: Provides historical data support for digital twin operation simulation and parameter configuration of new operation scenarios, and assists in the formulation of operation plans.

9. A protection method for an intelligent safety protection system for power operations, characterized in that, Includes the following steps: Step 1: System initialization and scene adaptation, each module performs self-test initialization, and the multi-scene adaptation module completes scene recognition, parameter configuration and device adaptation; Step 2: Multi-source fusion localization and 3D modeling, collecting multi-source localization data and image features, constructing a dynamic 3D model and correcting localization deviations; Step 3: Safety status perception and risk prediction, collect data from multiple sensors, identify explicit violations and predict implicit risks, and calculate the risk level; Step 4: Edge computing and digital twin risk assessment, local data processing, construction of digital twin model and simulation of work process to identify potential security risks; Step 5: Ensure 5G slicing communication transmission, transmit various types of data through 5G slicing, and monitor and adjust transmission quality; Step 6: Implement tiered intervention and control measures, and execute audible and visual warnings or emergency shutdown interventions based on the risk level, and record intervention feedback; Step 7: Data management and system optimization, storing and analyzing system operation data, optimizing system performance and archiving data; Step 8: The operation ends and the system is reset. All modules stop working and are reset to standby state.

10. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the system according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the system according to any one of claims 1-8.

Images