Multi-scene safety production supervision method, system and equipment based on intelligent positioning

Abstract

The invention relates to a multi-scene safety production supervision method, system and device based on intelligent positioning. The method comprises the steps that operation and maintenance monitoring is carried out on an application corresponding to each production scene through an enterprise side safety production risk monitoring and early warning subsystem, and production related information is acquired. Different source data in the production related information are integrally processed, a standard data set is generated, a label is added to each piece of source data in the standard data set, and a labeled data set is obtained. And constructing a safety production monitoring model according to the danger attributes, and inputting the labeled data set as a sample data set into the safety production monitoring model for optimization training to obtain an optimized safety production monitoring model. And deploying the optimized safety production monitoring model to the edge computing device in a distributed manner, so that the operation and maintenance monitoring subsystem carries out supervision processing according to the position information transmitted by the personnel positioning system and the production process information. By adopting the method, the real-time performance and accuracy of production safety monitoring in the dangerous chemical environment can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent production safety supervision, and in particular to a multi-scenario production safety supervision method, system and equipment based on intelligent positioning. Background Art

[0002] There are many potential safety hazards in the production and storage of hazardous chemicals, which not only threaten the lives of workers, but also may bring disastrous consequences to the company and the surrounding environment. Specifically, the following aspects are particularly prominent:

[0003] Hot work: Hot work ticket management involves major accident hazards. Hot work usually involves dangerous factors such as high temperature and flames. If it is not handled properly, it is very easy to cause fire or explosion accidents.

[0004] Liquid ammonia unloading: Liquid ammonia is a volatile and toxic chemical. Once it leaks during the unloading process, it will not only cause environmental pollution but also pose a serious threat to human health.

[0005] Key regulated processes: Certain hazardous chemical production processes have strict requirements on parameters such as temperature and pressure. Once the parameters deviate from the normal range, serious safety accidents may occur.

[0006] Abnormal operating conditions: Various abnormal operating conditions may occur during the production process, such as equipment failure, material leakage, etc. If these abnormal conditions are not handled in a timely manner, they may lead to production interruptions or even accidents.

[0007] In recent years, various safety accidents have frequently occurred in the production process of hazardous chemicals, causing huge economic losses and casualties. The production environment of hazardous chemicals is complex and changeable, and traditional safety supervision methods are difficult to achieve comprehensive coverage, and there are blind spots in supervision. Once a safety accident occurs, due to the lack of effective monitoring methods, it is often not possible to detect and take measures in time, resulting in the expansion of the accident. Operators work in a complex environment and face many safety risks, which require real-time monitoring and protection. The leakage or combustion of hazardous chemicals will produce toxic and harmful gases, causing pollution to the atmosphere, water bodies, etc., affecting the ecological environment and residents' health.

[0008] At present, traditional video surveillance can provide intuitive visual information and record the situation of the production site. However, there is passive monitoring: it relies on manual observation and cannot achieve 24-hour uninterrupted monitoring; the false alarm rate is high: the video surveillance system is prone to false alarms or missed alarms when identifying abnormal situations; the data processing capacity is limited: when faced with a large amount of video data, the processing efficiency is low. Therefore, although the use of the personnel positioning system alone can track the location of personnel in real time and improve the efficiency of emergency response, the accuracy is limited: in a complex environment, the positioning accuracy may be affected; the integration is not high: the use of the personnel positioning system alone cannot fully solve the safety supervision problem; data isolation: lack of linkage with other systems such as video surveillance, the data cannot be effectively integrated. In addition, the existing technology uses a comprehensive platform to integrate multiple monitoring methods and provide a unified management interface, but there are still compatibility issues between products of different manufacturers, resulting in unstable data transmission. With the increasing complexity and difficulty of the production process, the deployment and maintenance costs are high, and professional personnel are required to operate. When faced with massive data, the processing capacity is limited and instant feedback cannot be achieved. Summary of the invention

[0009] Based on this, it is necessary to provide a multi-scenario production safety supervision method, system and equipment based on intelligent positioning that can detect safety hazards in real time and improve the response efficiency of early warnings in response to the above technical problems.

[0010] A multi-scenario production safety supervision method based on intelligent positioning is applied to the production supervision Internet of Things, and the method includes:

[0011] According to the hazardous attributes of the production scenarios, the enterprise-side production safety risk monitoring and early warning subsystem conducts operation and maintenance monitoring for the scenario applications corresponding to each production scenario, and obtains production supervision-related information collected by camera equipment and personnel positioning systems.

[0012] Aggregate different source data in regulatory-related information to generate a standard data set, and add labels to each source data in the standard data set according to the preset annotation guidelines to obtain a labeled data set.

[0013] A production safety monitoring model is constructed according to the hazardous attributes, and the labeled data set is input into the production safety monitoring model as a sample data set for optimization training to obtain an optimized production safety monitoring model.

[0014] Distributed deployment has optimized the production safety monitoring model to the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0015] A multi-scenario production safety supervision system based on intelligent positioning, the system comprising:

[0016] The production-related information collection module is used to perform operation and maintenance monitoring for the scenario applications corresponding to each production scenario through the enterprise-side production safety risk monitoring and early warning subsystem according to the hazardous attributes of the production scenario, and to obtain production supervision-related information collected by the camera equipment and personnel positioning system.

[0017] The labeling module is used to aggregate different source data in the supervision-related information to generate a standard data set, and add labels to each source data in the standard data set according to the preset labeling guidelines to obtain a labeled data set.

[0018] The model training module is used to build a safe production monitoring model based on hazardous attributes, input the labeled data set as a sample data set into the safe production monitoring model for optimization training, and obtain an optimized safe production monitoring model.

[0019] The safety supervision module is used to distribute the optimized production safety monitoring model to the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0020] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0021] According to the hazardous attributes of the production scenarios, the enterprise-side production safety risk monitoring and early warning subsystem conducts operation and maintenance monitoring for the scenario applications corresponding to each production scenario, and obtains production supervision-related information collected by camera equipment and personnel positioning systems.

[0022] Aggregate different source data in regulatory-related information to generate a standard data set, and add labels to each source data in the standard data set according to the preset annotation guidelines to obtain a labeled data set.

[0023] A production safety monitoring model is constructed according to the hazardous attributes, and the labeled data set is input into the production safety monitoring model as a sample data set for optimization training to obtain an optimized production safety monitoring model.

[0024] Distributed deployment has optimized the production safety monitoring model to the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0025] The above-mentioned multi-scenario production safety supervision method, system and equipment based on intelligent positioning have the following technical effects compared with the existing technology: 1) Improve the real-time and accuracy of monitoring: Real-time monitoring: Through video AI technology and personnel positioning technology, real-time monitoring of the production site is realized, and safety hazards are discovered and handled in time. Personnel positioning: Using high-precision positioning technologies such as UWB and BLE to ensure the accuracy of personnel location information and improve management efficiency. 2) Reduce the load of the central server: Edge computing deployment: Assign computing tasks to edge devices to reduce the pressure on the central server and improve the response speed of the system. Data filtering: The edge device performs preliminary processing on the data and only uploads key information to reduce the data processing volume of the central server. 3) Enhance the intelligent decision-making support capability of the system: Intelligent analysis: By integrating video data and real-time data of the Internet of Things, comprehensive analysis is performed to improve the intelligent decision-making support capability of the system. Early warning mechanism: Automatically trigger the early warning mechanism, remind relevant personnel in various ways, and respond to abnormal situations in time. 4) Improve personnel safety management: Behavior recognition: Identify unsafe behaviors such as not wearing PPE and illegal operations, and improve the safety awareness of operators. Trajectory tracking: record the historical trajectory of personnel, analyze their behavior patterns, predict potential risk points, and strengthen safety management. 5) Improve emergency response capabilities: Rapid positioning: In an emergency, it is possible to quickly locate personnel in the affected area and speed up rescue. Plan provision: Provide emergency response plans to guide operators and managers to take correct actions in emergency situations. 6) Promote production efficiency and economic benefits: Reduce accidents: By timely discovering and handling safety hazards, reduce the possibility of accidents and reduce economic losses. Optimize management: Optimize production processes through data analysis, improve production efficiency, and bring higher economic benefits. 7) Support scenario expansion and future development: Modular design: The system adopts modular design to support the integration of future technologies, such as 5G communication and quantum computing. Flexible expansion: Supports rapid access to a variety of new scenarios, such as chemical warehouse management, transport vehicle monitoring, etc., to ensure the sustainable development of the system. 8) Promote the orderly transmission and processing of information. Step-by-step reporting mechanism: Through the step-by-step reporting mechanism, ensure that information is transmitted in an orderly manner according to the hierarchical structure and handle abnormal situations in a timely manner. Closed-loop management: Form a closed loop of information transmission to ensure that each piece of information has processing result feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is an application scenario diagram of a multi-scenario production safety supervision method based on intelligent positioning in an embodiment;

[0027] Figure 2 A flowchart of a multi-scenario production safety supervision method based on intelligent positioning in one embodiment;

[0028] Figure 3A flowchart of multi-scenario production safety supervision steps based on intelligent positioning in one embodiment;

[0029] Figure 4 It is a structural block diagram of a multi-scenario production safety supervision system based on intelligent positioning in one embodiment;

[0030] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0032] The multi-scenario production safety supervision method based on intelligent positioning provided by the present invention can be applied to Figure 1 In the application environment shown in the figure, the personnel positioning data comes from the personnel positioning system of the enterprise, and is directly connected to the interface of the enterprise personnel positioning system to complete the real-time upload of data. The main data of the personnel positioning system include: real-time positioning, estimated playback, electronic fence, etc. Through the interface, the data of the hazardous chemical enterprise production safety risk monitoring and early warning system of the Provincial Emergency Department is connected to obtain the equipment, process, access, and environmental detection data of major hazardous sources, double prevention and other systems, such as temperature, pressure, liquid level, etc. In order to make up for the problems of missing data and data omissions in the production safety risk monitoring and early warning system of hazardous chemical enterprises, the information of enterprise systems (such as DCS systems, SIS systems, etc.), sensor data (gas sensors, liquid level sensors, etc.), video data, etc. are accessed through the Internet of Things technology of the Internet of Things platform. The data is uploaded to the big data platform through the Internet of Things platform and the interface connection. The big data platform is responsible for processing, encrypting and real-time processing of massive data. Manual labeling is performed through the deep learning platform, and the relevant algorithm models in the AI ​​algorithm library are used on the platform for training and deployment. If an enterprise needs to generate a knowledge graph for subsequent retrieval as a knowledge base, it can create a knowledge graph through a knowledge graph engine. The creation process includes multiple links such as uploading data, knowledge extraction, knowledge mapping, graph production, and entity disambiguation. The model in the knowledge extraction process can be sent and deployed through the deep learning platform. After that, a model library of the "six properties" in different scenarios is formed through models in different fields, and applied in the upper-level application scenarios.

[0033] In one embodiment, Figure 2 As shown in the figure, a multi-scenario safety production supervision method based on intelligent positioning is provided. Figure 1 The IoT architecture in the example is used to illustrate the following steps:

[0034] Step 202, based on the hazardous attributes of the production scene, the enterprise-side production safety risk monitoring and early warning subsystem performs operation and maintenance monitoring for the scene application corresponding to each production scene, and obtains production supervision related information collected by the camera equipment and the personnel positioning system.

[0035] Step 204 , aggregate different source data in the supervision-related information to generate a standard data set, and add a label to each source data in the standard data set according to a preset labeling guide to obtain a labeled data set.

[0036] Step 206, constructing a production safety monitoring model according to the hazard attributes, inputting the labeled data set as a sample data set into the production safety monitoring model for optimization training, and obtaining an optimized production safety monitoring model.

[0037] Step 208, distributed deployment of the optimized production safety monitoring model on the edge computing device, so that the operation and maintenance monitoring subsystem performs supervision and processing according to the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0038] Compared with the existing technology, the above-mentioned multi-scenario production safety supervision method based on intelligent positioning has the following technical effects: 1) Improve the real-time and accuracy of monitoring: Real-time monitoring: Through video AI technology and personnel positioning technology, real-time monitoring of the production site is realized, and safety hazards are discovered and handled in time. Personnel positioning: Using high-precision positioning technologies such as UWB and BLE to ensure the accuracy of personnel location information and improve management efficiency. 2) Reduce the load of the central server: Edge computing deployment: Assign computing tasks to edge devices to reduce the pressure on the central server and improve the response speed of the system. Data filtering: The edge device performs preliminary processing on the data and only uploads key information to reduce the data processing volume of the central server. 3) Enhance the intelligent decision-making support capability of the system: Intelligent analysis: By integrating video data and real-time data of the Internet of Things, comprehensive analysis is performed to improve the intelligent decision-making support capability of the system. Early warning mechanism: Automatically trigger the early warning mechanism, remind relevant personnel in various ways, and respond to abnormal situations in time. 4) Improve personnel safety management: Behavior recognition: Identify unsafe behaviors such as not wearing PPE and illegal operations, and improve the safety awareness of operators. Trajectory tracking: Record the historical trajectory of personnel, analyze their behavior patterns, predict potential risk points, and strengthen safety management. 5) Improve emergency response capabilities: Rapid positioning: In an emergency, it is possible to quickly locate people in the affected area and speed up rescue. Plan provision: Provide emergency response plans to guide operators and managers to take correct actions in emergency situations. 6) Promote production efficiency and economic benefits: Reduce accidents: By timely discovering and handling safety hazards, reduce the possibility of accidents and reduce economic losses. Optimize management: Optimize production processes through data analysis, improve production efficiency, and bring higher economic benefits. 7) Support scenario expansion and future development: Modular design: The system adopts modular design to support the integration of future technologies, such as 5G communication and quantum computing. Flexible expansion: Supports rapid access to a variety of new scenarios, such as chemical warehouse management, transport vehicle monitoring, etc., to ensure the sustainable development of the system. 8) Promote the orderly transmission and processing of information. Step-by-step reporting mechanism: Through the step-by-step reporting mechanism, ensure that information is transmitted in an orderly manner according to the hierarchical structure and handle abnormal situations in a timely manner. Closed-loop management: Form a closed loop of information transmission to ensure that each piece of information has processing result feedback.

[0039] In one embodiment, if Figure 3As shown in the figure, firstly, the personnel positioning system is deployed according to the relevant requirements of the Ministry of Emergency Management, and the corresponding cameras are added according to the scene requirements, and the personnel positioning information and video data are uploaded to the big data platform through the local area network. The main data of the personnel positioning system include: real-time positioning, estimated playback, electronic fence, etc. Then, the data of the hazardous chemical enterprise production safety risk monitoring and early warning system of the Provincial Emergency Department is connected through the interface to obtain the equipment, process, access, and environmental detection data of major hazard sources, double prevention and other systems, such as temperature, pressure, liquid level, etc. In order to make up for the problems of data loss and data omission in the production safety risk monitoring and early warning system of hazardous chemical enterprises, the information of enterprise systems (such as DCS systems, SIS systems, etc.), sensor data (gas sensors, liquid level sensors, etc.), video history data, etc. are accessed through the Internet of Things technology of the Internet of Things platform. The data is uploaded to the big data platform through the Internet of Things platform and interface docking. The big data platform is responsible for processing, encrypting, fusing and real-time processing of massive data. Manual labeling is performed through the deep learning platform, and the relevant algorithm models in the AI ​​algorithm library are used on the platform for training and deployment. The trained model is sent to the edge measurement for deployment to provide decision-making and support for edge measurement. The edge-measured video and real-time positioning data are used by the model to determine whether hidden dangers have occurred in the scene. After checking, the on-duty personnel arrange on-site inspection personnel to check on the site. If the model prediction results are inconsistent with the inspection results, the warning information is a model misjudgment, and the information will enter the negative sample library for incremental training of the model. If the model prediction results are consistent with the on-site inspection results, the security inspectors need to take emergency safety measures quickly. If the processing is not completed within the specified time limit, the warning information will be reported step by step, and the company will be urged to complete the rectification as soon as possible through step-by-step supervision. After the rectification is completed within the time limit, the system will form a warning notification ledger, forming a closed loop of warning processing, and details can be viewed in the visual interface, such as processing time, processing personnel, cause of the accident, and accident handling method.

[0040] In one of the embodiments, after setting the warning rules according to the dangerous attributes of the production scene, the enterprise-side production safety risk monitoring and early warning subsystem performs operation and maintenance monitoring for the data characteristics of the scene application corresponding to each production scene, and obtains the production supervision related information collected by the camera equipment and the personnel positioning system. The production supervision related information includes: personnel positioning system data and monitoring camera data. The personnel positioning system data includes: personnel location information, personnel status information, personnel safety alarm information and emergency information. The monitoring camera data includes: video image data, personnel behavior data, equipment status data, environmental data and monitoring safety alarm information.

[0041] It is worth mentioning that the personnel positioning system data is as follows: 1. Personnel location information includes: Real-time location: Real-time acquisition of the specific location of employees in the factory, including floors, areas, etc. Historical trajectory: Record the historical movement trajectory of employees for backtracking and analysis. 2. Personnel status information includes: Work status: Employee work status (such as work, rest, training, etc.). Activity time: Employees' stay time and activity frequency in different areas. Abnormal behavior: Detect whether employees have abnormal behavior, such as staying in dangerous areas for a long time. 3. Safety alarm information includes: Entering restricted areas: Detect whether employees enter prohibited areas. Overtime stay: Detect whether employees stay in a certain area for more than the specified time. Not wearing positioning devices: Detect whether there are employees who are not wearing positioning devices. 4. Emergency information includes: SOS alarm: When employees encounter an emergency, they can send an SOS alarm signal through the positioning device. Evacuation path: In an emergency, provide employees with the optimal evacuation path.

[0042] The data obtained by the surveillance camera are as follows: 1. Video image data includes: Real-time video: Real-time transmission of video images of the monitoring area for real-time monitoring. Historical video: Store historical video for subsequent tracing and analysis. 2. Personnel behavior data includes: Behavior recognition: Use AI video analysis technology to identify employee behavior, such as whether they wear a helmet, whether they comply with operating procedures, etc. Abnormal behavior: Detect whether employees have abnormal behavior, such as running, falling, gathering, etc. 3. Equipment status data includes: Equipment operation status: Monitor the operation status of the equipment, such as whether the equipment is operating normally, whether there is abnormal vibration, etc. Equipment maintenance: Record the maintenance of the equipment to ensure that the equipment is in good condition. 4. Environmental data includes: Environmental parameters: Monitor environmental parameters such as temperature, humidity, and concentration of toxic and harmful gases through the built-in sensors of the camera. Environmental changes: Detect changes in the environment, such as smoke, flames, leaks, etc. 5. Security alarm information includes: Intrusion detection: Detect whether unauthorized personnel enter the monitoring area. Abnormal activity: Detect whether there are abnormal activities in the monitoring area, such as abnormal sounds, abnormal lighting, etc.

[0043] In one of the embodiments, a labeling guide is set based on surveillance camera data; the labeling guide includes: video image data labeling, personnel behavior data labeling, equipment status data labeling, environmental data labeling, and security alarm information labeling. Video image data labeling is used to label objects detected in the video, video boundary boxes, and object key points. Personnel behavior data labeling is used to label personnel safety behaviors and personnel abnormal behaviors identified by surveillance cameras. Equipment status data labeling is used to label equipment operation status and equipment maintenance status. Environmental data labeling is used to label environmental parameters and environmental changes. Security alarm information labeling is used to label unauthorized intrusion detection and abnormal activities in the monitoring area.

[0044] It is worth noting that the application scenarios of hazardous chemical production include the following Table 1 and Table 2:

[0045] Table 1 Liquid ammonia filling scenarios

[0046]

[0047] Table 2 Hot Work

[0048]

[0049] In one of the embodiments, a production safety monitoring model is constructed based on hazardous attributes, and the labeled data set is input as a sample data set into the production safety monitoring model for monitoring. If the hazardous attribute is a key supervision process, the real-time production data of hazardous chemical enterprises is monitored for process anomalies through the interface of the supervision process early warning model, and a process elimination strategy is generated. If the hazardous attribute is an abnormal working condition, an AI model is used to process the safety alarm information in the scenario application, and the process of safety alarm information is monitored through the abnormal working condition monitoring model to generate an abnormal early warning strategy. The process elimination strategy and the abnormal early warning strategy are input into the production safety monitoring model for optimization training to obtain an optimized production safety monitoring model.

[0050] It is worth noting that for key regulated processes, the real-time production data of hazardous chemical enterprises (DCS system data, GDS system data, SIS system data, etc.) are connected through interfaces to monitor the relevant alarm data of the process. The early warning model for key regulated processes ensures that there are no abnormal alarms in the process, and determines whether the enterprise uses processes and equipment listed in the catalog of obsolete safety technology processes and equipment (major safety hazards). For abnormal working conditions, AI models are used to see scene models such as wearing protective equipment and gathering of people, and then toxic and harmful gas alarms are connected through the Internet of Things to monitor toxic and harmful gas data in real time. Through the abnormal working condition monitoring model, it is determined whether the enterprise has equipment involving key regulated hazardous chemical processes that has not achieved automated control, the system has not achieved the emergency stop function, and the equipment's automated control system and emergency stop system have not been put into use (major safety hazards).

[0051] In one embodiment, the optimized production safety monitoring model is distributedly deployed on the edge computing device so that the operation and maintenance monitoring subsystem performs supervision according to the received location information transmitted by the personnel positioning system, the production process information transmitted by the camera device, and the preset early warning strategy. The specific steps are as follows: 1. Video data: Use a computer vision model based on deep learning (such as YOLO, Faster R-CNN, etc.) to analyze the video stream in real time to identify unsafe behaviors, equipment abnormalities, etc. 2. Personnel positioning data: Receive and parse personnel positioning information to determine the specific location of each operator. 3. Comprehensive data analysis: Combine the video analysis results with the personnel positioning information, analyze the fused data through machine learning algorithms, identify abnormal situations, provide decision support, and identify potential safety hazards. 4. Early warning strategy: The model makes decisions and determines whether a hidden danger occurs. If a hidden danger occurs, the safety duty personnel immediately arranges an on-site inspection. If the inspection result is consistent with the model discrimination result, the corresponding safety measures should be taken; if not, it is a model misjudgment, and the sample is re-labeled and learned as a negative sample. If the rectification is completed within the specified time limit, the alarm will be eliminated. If the rectification is not completed within the specified time limit, the implementation of safety measures will be supervised and early warning will be issued through the step-by-step reporting mechanism. 5. Reporting mechanism: If the rectification is not completed within a certain time limit, it will be reported step by step through the company structure from the workshop, safety and environmental protection department, company management, company legal person, district and county emergency management department, municipal emergency management department and provincial emergency management department, so that the company can take emergency plans in time to prevent the situation from spreading.

[0052] In one embodiment, the early warning strategy is as follows: if the inspection result is consistent with the model discrimination result, the safety measures corresponding to the safety hazard are taken. If they are inconsistent, it is a model misjudgment, and the current data sample is re-labeled and the monitoring model is learned as a negative sample; if the rectification is completed within the specified period, the alarm is eliminated; if the rectification is not completed within the specified period, the implementation of safety measures is supervised and an early warning is issued through the step-by-step reporting mechanism; if the rectification is not completed within a certain time limit through the step-by-step reporting mechanism, the problem is reported step by step through the company structure from the workshop, the safety and environmental protection department, the company management, the company legal person, the district and county emergency management department, the municipal emergency management department and the provincial emergency management department.

[0053] In one of the embodiments, the operation and maintenance monitoring subsystem displays real-time monitoring images, personnel location distribution maps, and warning information through a Web or mobile application, so that administrators can query historical monitoring records for data analysis and backtracking, and security duty personnel can cancel alarms through a visual interface.

[0054] In one of the embodiments, after the model is sent to the edge measurement, the edge measurement video and real-time positioning data are used to determine whether hidden dangers in the scene occur. After checking, the on-duty personnel arrange on-site inspection personnel to check on site. If the model prediction result is inconsistent with the inspection result, the warning information is a model misjudgment, and the information will enter the negative sample library for incremental training of the model. If the model prediction result is consistent with the on-site inspection result, the security inspector needs to take emergency safety measures quickly. If it is not handled within the specified time limit, the warning information will be reported step by step, and the enterprise will be urged to complete the rectification as soon as possible through step-by-step supervision. After the rectification is completed within the time limit, the system forms a warning notification ledger, forming a closed loop of warning processing, and details can be viewed in the visual interface, such as processing time, handler, cause of the accident, and accident handling method.

[0055] It should be understood that although Figure 2-3 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2-3 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0056] In one embodiment, Figure 4 As shown, a multi-scenario production safety supervision system based on intelligent positioning is provided, including: a production related information collection module 402, a labeling module 404, a model training module 406 and a safety supervision module 408, wherein:

[0057] The production-related information collection module 402 is used to perform operation and maintenance monitoring for the scene application corresponding to each production scene through the enterprise-side production safety risk monitoring and early warning subsystem according to the dangerous attributes of the production scene, and obtain production supervision-related information collected by the camera equipment and the personnel positioning system;

[0058] The labeling module 404 is used to aggregate different source data in the supervision-related information to generate a standard data set, and add a label to each source data in the standard data set according to a preset labeling guide to obtain a labeled data set.

[0059] The model training module 406 is used to construct a production safety monitoring model according to the hazard attributes, and input the labeled data set as a sample data set into the production safety monitoring model for optimization training to obtain an optimized production safety monitoring model.

[0060] The safety supervision module 408 is used for distributed deployment of the optimized production safety monitoring model on the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0061] For the specific limitations of the multi-scenario production safety supervision system based on intelligent positioning, please refer to the limitations of the multi-scenario production safety supervision method based on intelligent positioning above, which will not be repeated here. Each module in the above-mentioned multi-scenario production safety supervision system based on intelligent positioning can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0062] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input system connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a multi-scenario safety production supervision method based on intelligent positioning is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input system of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0063] Those skilled in the art will understand that Figure 4-5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0064] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0065] According to the hazardous attributes of the production scenarios, the enterprise-side production safety risk monitoring and early warning subsystem conducts operation and maintenance monitoring for the scenario applications corresponding to each production scenario, and obtains production supervision-related information collected by camera equipment and personnel positioning systems.

[0066] Aggregate different source data in regulatory-related information to generate a standard data set, and add labels to each source data in the standard data set according to the preset annotation guidelines to obtain a labeled data set.

[0067] A production safety monitoring model is constructed according to the hazardous attributes, and the labeled data set is input into the production safety monitoring model as a sample data set for optimization training to obtain an optimized production safety monitoring model.

[0068] Distributed deployment has optimized the production safety monitoring model to the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

[0069] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention 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. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0070] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0071] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A multi-scenario safety production supervision method based on intelligent positioning, characterized in that: Applied to production supervision Internet of Things, the method includes: According to the dangerous attributes of the production scene, the enterprise-side production safety risk monitoring and early warning subsystem conducts operation and maintenance monitoring for the scene application corresponding to each production scene, and obtains production supervision related information collected by the camera equipment and personnel positioning system; Aggregate different source data in the supervision-related information to generate a standard data set, and add labels to each source data in the standard data set according to a preset labeling guide to obtain a labeled data set; Constructing a production safety monitoring model according to the hazardous attributes, inputting the labeled data set as a sample data set into the production safety monitoring model for optimization training, and obtaining an optimized production safety monitoring model; The optimized production safety monitoring model is distributedly deployed on edge computing devices so that the operation and maintenance monitoring subsystem performs supervision and processing based on the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment, and the preset early warning strategy.

2. The method according to claim 1, characterized in that According to the dangerous attributes of the production scene, the enterprise-side production safety risk monitoring and early warning subsystem performs operation and maintenance monitoring for the scene application corresponding to each production scene, and obtains production supervision-related information collected by the camera equipment and personnel positioning system, including: After setting the warning rules according to the dangerous attributes of the production scene, the enterprise-side production safety risk monitoring and early warning subsystem performs operation and maintenance monitoring on the data features of the scene application corresponding to each production scene, and obtains the production supervision-related information collected by the camera equipment and the personnel positioning system; The production supervision related information includes: personnel positioning system data and surveillance camera data; The personnel positioning system data includes: personnel location information, personnel status information, personnel safety alarm information and emergency information; The monitoring camera data includes: video image data, personnel behavior data, equipment status data, environmental data and monitoring security alarm information.

3. The method according to claim 2, characterized in that Setting a labeling guide according to the surveillance camera data; the labeling guide includes: video image data labeling, personnel behavior data labeling, equipment status data labeling, environmental data labeling and security alarm information labeling; The video image data annotation is used to annotate objects detected in the video, video bounding boxes, and object key points; The personnel behavior data annotation is used to annotate the personnel safety behavior and personnel abnormal behavior identified by the surveillance camera; The device status data annotation is used to annotate the device operation status and the device maintenance status; The environmental data annotation is used to annotate environmental parameters and environmental changes; The security alarm information marking is used to mark unauthorized intrusion detection and abnormal activities in the monitoring area.

4. The method according to claim 3, characterized in that: Constructing a production safety monitoring model according to the hazardous attributes, inputting the labeled data set as a sample data set into the production safety monitoring model for optimization training, and obtaining an optimized production safety monitoring model, including: A production safety monitoring model is constructed according to the hazardous attribute, and the labeled data set is input as a sample data set into the production safety monitoring model for monitoring. If the hazardous attribute is a key supervision process, the real-time production data of the hazardous chemical enterprise is monitored for process anomalies through the interface of the supervision process early warning model, and a process elimination strategy is generated; if the hazardous attribute is an abnormal working condition, an AI model is used to process the safety alarm information in the scenario application, and the safety alarm information is monitored through the abnormal working condition monitoring model to generate an abnormal early warning strategy; The process elimination strategy and the abnormal warning strategy are input into the safe production monitoring model for optimization training to obtain an optimized safe production monitoring model.

5. The method according to claims 1 to 4, characterized in that The specific early warning strategy is: if the inspection result is consistent with the model discrimination result, then take safety measures corresponding to the safety hazard; if they are inconsistent, it is a model misjudgment, and the current data sample is used as a negative sample to re-label and learn the monitoring model; If the rectification is completed within the specified period, the alarm will be eliminated; If the rectification is not completed within the specified time limit, the implementation of safety measures will be supervised and early warnings will be issued through a step-by-step reporting mechanism; if the rectification is not completed within a certain time limit through the step-by-step reporting mechanism, the problem will be reported step by step through the company structure, from the workshop, safety and environmental protection department, company management, company legal person, district and county emergency management department, municipal emergency management department and provincial emergency management department.

6. The method according to claims 1 to 4, characterized in that The operation and maintenance monitoring subsystem displays real-time monitoring images, personnel location distribution maps and warning information through the Web or mobile applications, so that administrators can query historical monitoring records for data analysis and backtracking, and security duty personnel can cancel alarms through a visual interface.

7. A multi-scenario production safety supervision system based on intelligent positioning, characterized in that: The system comprises: The production-related information collection module is used to perform operation and maintenance monitoring for the scene application corresponding to each production scene through the enterprise-side production safety risk monitoring and early warning subsystem according to the dangerous attributes of the production scene, and obtain production supervision-related information collected by the camera equipment and personnel positioning system; A labeling module is used to aggregate different source data in the supervision-related information to generate a standard data set, and add a label to each source data in the standard data set according to a preset labeling guide to obtain a labeled data set; A model training module, used to construct a production safety monitoring model according to the hazardous attributes, input the labeled data set as a sample data set into the production safety monitoring model for optimization training, and obtain an optimized production safety monitoring model; The safety supervision module is used to distribute the optimized production safety monitoring model to the edge computing device, so that the operation and maintenance monitoring subsystem can perform supervision and processing according to the location information transmitted by the personnel positioning system, the production process information transmitted by the camera equipment and the preset early warning strategy.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

Images