Safety monitoring and management methods, systems and media for chemical product production workshops

Abstract

This invention discloses a method, system, and medium for safety monitoring and management in chemical product production workshops, relating to the field of workshop safety management technology. The method includes: dividing the chemical product production workshop according to regional functions and establishing a regional transmission topology diagram; deploying multiple types of monitors to collect multi-source detection data; training a classifier to perform safety identification and classification, and locating the safety identification results; based on the regional transmission topology diagram, performing regional point identification and regional transmission risk prediction, generating early warning levels, and implementing graded early warning and emergency management control of the product production process based on the early warning levels. This invention solves the technical problems of incomplete and inaccurate safety monitoring and identification in chemical product production workshops, making it difficult to effectively predict risks, leading to a high probability of accidents and difficulty in ensuring production safety. It achieves graded early warning and emergency management control in the management of chemical product production processes, improving the technical effect of workshop safety monitoring and management.

Description

Technical Field

[0001] This invention relates to the field of workshop safety management technology, specifically to safety monitoring and management methods, systems, and media for chemical product production workshops. Background Technology

[0002] In the production of chemical products, numerous flammable, explosive, toxic, and hazardous chemicals are involved, along with complex processes such as high temperature and high pressure, creating many potential hazards in production workshops. Current technologies for safety monitoring and management in chemical workshops largely rely on manual inspections and single-type sensors. This not only suffers from limited monitoring range and incomplete data collection but also makes it difficult to accurately identify and analyze abnormal hazards in complex environments in real time. Furthermore, the lack of a systematic risk transmission and prediction mechanism makes it impossible to anticipate the impact of risks on surrounding areas, leading to difficulties in timely response when accidents occur and seriously threatening production safety and personnel safety.

[0003] Existing technologies suffer from incomplete and inaccurate safety monitoring and identification in chemical product production workshops, making it difficult to effectively predict risks and resulting in a high probability of accidents and difficulty in ensuring production safety. Summary of the Invention

[0004] This application provides a method, system, and medium for safety monitoring and management in chemical product production workshops, which addresses the technical problems in the prior art where safety monitoring in chemical product production workshops is incomplete, identification is inaccurate, and risks are difficult to predict effectively, resulting in a high probability of accidents and difficulty in ensuring production safety.

[0005] In view of the above problems, this application provides a method, system and medium for safety monitoring and management in chemical product production workshops.

[0006] A first aspect of this application provides a safety monitoring and management method for a chemical product production workshop, the method comprising:

[0007] A regional transmission topology diagram is established for the chemical product production workshop according to its functional area division. Multiple types of monitors are deployed at functional area nodes such as reaction units, raw material storage areas, high-temperature and high-pressure equipment areas, and power auxiliary facilities to collect multi-source detection data. A classifier is trained based on safety level classification rules, and safety identification and classification are performed using the classifier based on the multi-source detection data to locate the safety identification results. Based on the regional transmission topology diagram and the safety identification results, regional point identification and regional transmission risk prediction are performed, generating early warning levels. Based on the early warning levels, graded early warnings and emergency management and control of the product production process are implemented.

[0008] A second aspect of this application provides a safety monitoring and management system for a chemical product production workshop, the system comprising:

[0009] The module includes a topology map creation module for establishing a regional transmission topology map of the chemical product production workshop according to regional functions; a multi-source detection data acquisition module for deploying various types of monitors at functional area nodes such as reaction units, raw material storage areas, high-temperature and high-pressure equipment areas, and power auxiliary facilities to collect multi-source detection data; a safety identification result location module for training a classifier based on safety level classification rules, performing safety identification and classification based on the multi-source detection data, and locating the safety identification result; and a warning level generation module for performing regional point identification and regional transmission risk prediction based on the regional transmission topology map and the safety identification result location, generating a warning level, and performing graded warnings and emergency management control of the product production process based on the warning level.

[0010] In a third aspect of this application, a computer-readable storage medium is provided, storing a computer program for executing the safety monitoring and management method for a chemical product production workshop provided in this application.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] A regional transmission topology map was established for the chemical product production workshop according to its functional areas. Multiple types of monitors were deployed to collect multi-source detection data. A classifier was trained, and safety identification and classification were performed based on the multi-source detection data to pinpoint the safety identification results. Based on the regional transmission topology map and the safety identification results, regional point identification and regional transmission risk prediction were performed, generating early warning levels. Based on these early warning levels, graded early warning and emergency management control of the product production process were implemented. This achieved the technical effect of graded early warning and emergency management control in the chemical product production process management, improving the workshop's safety monitoring and management level. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart illustrating a safety monitoring and management method for a chemical product production workshop provided in an embodiment of this application.

[0015] Figure 2 A schematic diagram of the structure of the safety monitoring and management system for a chemical product production workshop provided in this application embodiment.

[0016] Figure labeling: Topology diagram creation module 10, multi-source detection data acquisition module 20, safety identification result location module 30, early warning level generation module 40. Detailed Implementation

[0017] This application provides a method, system, and medium for safety monitoring and management in chemical product production workshops, which addresses the technical problems in existing technologies such as incomplete and inaccurate safety monitoring in chemical product production workshops, difficulty in effectively predicting risks, resulting in a high probability of accidents and difficulty in ensuring production safety.

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] Example 1, as Figure 1 As shown, this application provides a safety monitoring and management method for chemical product production workshops, the method comprising:

[0020] Step S100: Divide the chemical product production workshop according to regional functions and establish a regional transmission topology diagram.

[0021] Specifically, the workshop is functionally divided into zones based on a comprehensive consideration of various factors, including: process flow (analyzing the processing sequence from raw material input to finished product output); equipment interrelationships (clarifying the collaborative relationships and energy transfer paths between different pieces of equipment); spatial location (determining the actual location of each zone based on the workshop's architectural layout); material properties (analyzing the flammability, explosiveness, toxicity, and other characteristics of materials stored or used in each zone); and personnel and energy flow (understanding the daily activity routes of personnel and the direction of energy transmission). After the division, each functional zone is defined as a node, assigned node attributes containing basic information such as node number, functional type, and material or energy risk level. Then, directed connections are established between nodes according to process flow, equipment interrelationships, and spatial location, with detailed records of the source node, target node, connection type, material hazard attributes, and hazardous material transmission probability for each edge. Using this data, the transmission risk weight of each edge is calculated and assigned, thereby constructing a weighted directed graph to form a regional transmission topology diagram. Simultaneously, based on the real-time process status, the risk level of process response will be dynamically analyzed, regional risk adjustment boundaries will be configured, functional partition boundaries will be reconstructed, the weight of adjacent edges will be adjusted according to the span of the boundary reconstruction area, and the regional transmission topology diagram will be dynamically updated in a time-series manner to accurately reflect the real-time status of the workshop. This will provide an accurate and effective infrastructure for subsequent multi-source data collection, safety identification and classification, risk prediction, early warning and emergency management based on this diagram.

[0022] Step S200: Install multiple types of monitors at functional nodes of the reaction unit, raw material storage area, high temperature and high pressure equipment area, and power auxiliary facilities to collect multi-source detection data.

[0023] Specifically, in the reaction unit, due to the complexity of the chemical reaction process and the potential for generating flammable gases, toxic substances, and heat changes, combustible gas detectors, toxic gas detectors, and temperature sensors are strategically deployed according to specific process requirements. Combustible gas detectors monitor for leaks in real time to prevent fires and explosions; toxic gas detectors protect personnel from toxic substances; and temperature sensors promptly detect temperature anomalies during the reaction process to prevent runaway reactions.

[0024] In the raw material storage area, considering the hazards of the stored materials, pressure sensors are installed in addition to flammable and toxic gas detectors. This is because some raw materials may experience pressure changes during storage due to factors such as temperature and chemical reactions; the pressure sensors can monitor pressure values ​​in real time to ensure storage safety.

[0025] High-temperature and high-pressure equipment areas are prone to accidents, making vibration sensors crucial. They monitor equipment vibration and analyze the data to identify potential malfunctions. Temperature and pressure sensors are also essential for monitoring temperature and pressure during equipment operation, ensuring stable operation.

[0026] Power auxiliary facilities provide power support for the entire workshop, and monitoring their operational status is equally important. Video monitors provide a direct view of the equipment's operation, facilitating the timely detection of anomalies. In addition, temperature and pressure sensors are also installed to ensure that the power auxiliary facilities operate within safe parameter ranges.

[0027] By deploying various types of monitors at these functional area nodes, multi-source detection data such as combustible gas concentration, toxic substance leakage, temperature, pressure, and equipment vibration can be collected. This data not only provides rich information for subsequent training of classifiers based on safety level classification rules, but also serves as an important basis for regional point identification, regional transmission risk prediction, and the generation of early warning levels for the implementation of tiered early warning and emergency management, thus comprehensively ensuring the safe production of chemical product manufacturing workshops.

[0028] Step S300: Train a classifier based on security level classification rules, perform security identification and classification through the classifier based on the multi-source detection data, and locate the security identification result.

[0029] Specifically, historical accident records are collected, and safety grading rules are established with reference to industry standards. Based on the safety sample characteristics, types, and corresponding safety level labels in the historical accident records, datasets for training and validation are constructed. The training dataset is used to train the selected classifier, while the validation dataset is used to evaluate the classifier's performance and ensure its accuracy. During this process, if new monitored hazardous information is encountered (i.e., situations where the safety type differs from that in the historical accident records), hazard characteristics are collected based on this new information, and the corresponding safety type and safety level are labeled. Incremental training and validation sets are constructed, and the initially trained classifier is then incrementally trained, continuously optimizing the classifier until the validation convergence target is reached, thereby obtaining a high-performance classifier capable of accurately identifying various hazards. After classifier training is complete, multi-source detection data collected from nodes in the reaction unit, raw material storage area, high-temperature and high-pressure equipment area, and power auxiliary facility functional area are input into the classifier. Based on learned rules and patterns, the classifier performs in-depth analysis and processing of the data to quickly determine the type of safety, such as flammable gas leaks, toxic gas leaks, or abnormalities caused by equipment malfunctions, and determines its safety level, accurately locating the safety identification results. These results will serve as a key basis for subsequent regional point identification, regional transmission risk prediction, and the generation of early warning levels for tiered early warning and emergency management, providing strong support for ensuring the safety of chemical production workshops.

[0030] Step S400: Based on the regional transmission topology diagram and the safety identification results, perform regional point identification and regional transmission risk prediction, generate early warning levels, and perform graded early warning and emergency management control of the product production process based on the early warning levels.

[0031] Specifically, when performing regional point identification and regional transmission risk prediction, the initial risk nodes are first mapped to a dynamic regional transmission topology diagram. For dynamic transmission weight calculation, based on real-time process parameters (pressure change rate, material flow rate) and environmental parameters (wind speed, obstacle density), the formula is used... Adjust the propagation weights of the connecting edges. Where w ij Representing the initial edge weight, it is the basic weight value for risk propagation from node i to node j without considering real-time parameter changes; ΔP ij The pressure gradient between adjacent nodes reflects the changes in pressure between them. The larger the pressure gradient, the greater its impact on the risk transmission weight. wind This is real-time wind speed, reflecting the current wind conditions in the environment. Wind speed affects the diffusion of hazardous substances, and thus influences the transmission weight; v refThe reference wind speed serves as a benchmark for comparison, used to measure the influence of real-time wind speed on the conduction weight; α and β are weighting coefficients used to adjust the influence of pressure gradient and wind speed on the conduction weight; w' ij (t) represents the connection edge propagation weight at time t after real-time parameter correction, comprehensively reflecting the weight of risk propagation between nodes under the current operating conditions. In the risk diffusion simulation, starting from the initial risk node, an improved graph propagation algorithm combining Monte Carlo random walks is used, through formula R... j (t)=∑ i∈邻接节点 R i (t-1)·w' ij (t)·λ ij (t) Predict the risk transmission path and risk value. Where R... i (t-1) represents the risk value of adjacent node i at time t-1, which is the basic data for calculating the risk value of the current node; w' ij (t) represents the corrected transmission weight from node i to node j calculated earlier, which determines the proportion of risk transmitted from node i to node j; λ ij (t) is the dynamic attenuation factor, which is inversely proportional to the obstacle distribution density. The denser the obstacle distribution, the greater the attenuation factor. ij The smaller the (t) value, the more the risk is hindered and attenuated during transmission; R j (t) is the risk value calculated by the target node j at time t, which is obtained by weighted summation of the relevant data of the adjacent nodes.

[0032] Simultaneously, risk center points are located and determined based on safety identification results, completing regional point identification and classifying point risk levels. Combining the regional point identification risk level with the regional transmission risk prediction results, a warning level is generated based on a preset fuzzy logic rule table. When the warning level is red (global high risk), the automated control function of the distributed control system (DCS) is used to cut off the valves of material supply pipelines in the hazardous area and close the valves of related equipment; the entire area spray system and inert gas injection device are activated through the industrial automation system; alternative production paths are generated based on the simulation capabilities of the digital twin model; intelligent algorithms are used to adjust the material conveying system to bypass the high-risk area; reinforcement learning algorithms are used to optimize the equipment control system, ensuring that equipment in non-hazardous areas starts and stops according to the new sequence, maintaining more than 50% of production capacity. If the warning level is orange (regional medium risk), the hazardous area is isolated using physical barriers; the reaction temperature, flow rate, and other parameters of upstream and downstream process equipment are adjusted through the industrial parameter adjustment system; relying on the data processing and real-time control capabilities of edge computing nodes, equipment operating parameters are collected and optimized; and an explosion-proof robot is directed to enter the hazardous area to perform tasks through a robot scheduling system. For yellow or blue alerts (local low risk), the monitoring frequency is increased to once per second by upgrading sensors and monitoring systems, and warnings are issued using audible and visual alarm devices; information is pushed to inspection personnel using the visualization and positioning functions of AR devices; based on the results of the risk transmission prediction model, emergency resources are deployed to high-probability risk areas in advance through logistics scheduling, ultimately achieving accurate graded early warning and emergency management and control of the product production process.

[0033] In one possible implementation, step S200 further includes:

[0034] Step S210: The multi-type monitors include: combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors, wherein the monitor deployment nodes in the functional areas include one or more of the combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors.

[0035] Specifically, to comprehensively and accurately monitor the workshop's operational status, a variety of monitoring equipment was selected, including combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video surveillance. Different functional areas face varying risks due to their different characteristics, resulting in different detector deployments. In the raw material storage area, where the stored materials are mostly flammable, explosive, or toxic substances, combustible gas detectors and toxic gas detectors are deployed to monitor for leaks in real time. Temperature and pressure sensors are also installed as needed, as changes in temperature and pressure can affect the stability of raw materials and even lead to safety accidents. In the high-temperature and high-pressure equipment area, vibration sensors are crucial. They can detect potential equipment malfunctions by monitoring vibration amplitude and frequency. Temperature and pressure sensors are also essential to prevent dangers caused by overheating or overpressure. In the reaction unit, due to the complexity of chemical reactions, in addition to combustible and toxic gas detectors, temperature changes during the reaction process must be closely monitored; therefore, temperature sensors are typically standard monitoring equipment in this area. In the power auxiliary facilities area, video monitors provide a clear view of the equipment's operating status, allowing operators to monitor the situation at any time. Temperature and pressure sensors are used to monitor key parameters during equipment operation, ensuring stable operation. In short, each functional area will select one or more of these various types of monitors for appropriate deployment based on its own risk characteristics, thereby achieving comprehensive and effective monitoring of the chemical product production workshop and providing strong support for safe production.

[0036] In one possible implementation, step S100 further includes:

[0037] Step S110: Analyze the characteristic relationships of each functional area according to the process flow, equipment correlation, spatial location relationship, material properties, personnel and energy flow.

[0038] Step S120: Define each functional area as a node. The node attributes include node number, functional type, and basic information on the material or energy risk level.

[0039] Step S130: Establish directed connections between nodes according to the process flow, equipment correlation, and spatial location relationship. Each edge includes the source node and the target node, connection type, material hazard attributes, and probability of hazardous material transmission.

[0040] Step S140: Calculate the transmission risk weight of each edge according to the safety level of the material in the connecting edge and the probability of transmission of hazardous material, assign weights to the edges, establish a weighted directed graph based on the nodes, edges and edge weights, and obtain the transmission topology graph of the region.

[0041] Specifically, the process begins with an in-depth study of the technological flow, meticulously analyzing each step from raw material input to product output. This clarifies the sequence and role of each functional area within the entire production chain, as well as the flow paths and transformation processes of materials between different areas. Simultaneously, based on equipment interrelationships, the collaborative operation relationships between equipment within each area are analyzed to determine which equipment has connections for material transport, energy transfer, or signal interaction, thereby identifying the connections between functional areas arising from equipment collaboration. From a spatial perspective, the actual distribution of each functional area within the workshop is precisely assessed, including the relative positions, distances, and layout characteristics between areas, as these factors affect the spread and speed of hazardous substances. Regarding material properties, the chemical characteristics of the substances involved in each area are carefully analyzed, such as whether they possess flammable, explosive, toxic, harmful, or highly corrosive properties, as well as their stability and reactivity under different conditions. Furthermore, close attention is paid to personnel and energy flow, with detailed records kept of personnel's daily activity trajectories, operating procedures, and frequently visited areas. The direction of energy transmission, transformation forms, and concentration areas within the workshop are also analyzed. Through a comprehensive analysis of the process flow, equipment correlation, spatial relationship, material properties, personnel and energy flow, we can deeply explore the internal connections and interactions between various functional areas, laying a solid foundation for the subsequent establishment of a regional transmission topology diagram.

[0042] Based on the comprehensive analysis of each area in the workshop, each functional area was defined as a separate node, aiming to provide a clear and independent research object for subsequent risk analysis and management. Each node was assigned specific attribute information, with the node number serving as a unique identifier. This allows for quick and accurate location and differentiation of different functional areas during data processing and subsequent operations. Taking a chemical product production workshop as an example, each node was assigned specific attribute information: Node number: using the rule of "first letter of area type - sequential number," such as R-001 for the raw material storage area and P-003 for the reaction unit. This serves as a unique identifier for the node, facilitating rapid retrieval of relevant parameters for the ethanol storage area in the data management system using R-001. The functional type clearly defines the specific task that area undertakes in the chemical production process, such as whether it is a raw material storage area responsible for storing raw materials, a reaction unit conducting chemical reactions, or a power auxiliary facility area providing power support for the entire production process. This helps to grasp the nature and role of each area from a macro perspective. The basic information on the risk level of a substance or energy source is a preliminary risk assessment based on a comprehensive evaluation of the properties of substances within the area (such as flammability, explosiveness, toxicity, etc.) and the characteristics of energy flow (such as the presence of high-temperature and high-pressure energy sources). Taking a chemical product production workshop as an example, in the raw material storage area (R-001): ethanol is a flammable liquid (flash point 12℃, risk level M), and chlorine is a highly toxic gas (LC50 = 90 mg / m³). 3The risk level of this node is H, based on a comprehensive assessment. The reaction unit (P-003) is at a high temperature (200℃) with a risk of chlorine leakage; its energy risk level is M, its material risk level is H, and its overall basic risk level is H. The power auxiliary facility (A-005) is a high-pressure steam (10MPa) source of energy hazard (risk level M), but does not store highly toxic substances; its overall basic risk level is M. This information provides the foundational data for subsequent risk transmission analysis based on the topology, making the analysis of the overall workshop risk situation more targeted and accurate.

[0043] Based on the process flow, the connection sequence between nodes is determined along the material flow path, such as from the raw material storage area node to the reaction unit node, reflecting the flow of material input into production. Simultaneously, considering equipment interrelationships, the relationships between equipment, such as energy transfer and material transport, are analyzed, and connections are established between corresponding nodes. If two pieces of equipment are located in different functional areas but collaborate closely, a connection edge is constructed between the nodes representing these two areas. When considering spatial relationships, adjacent areas or areas that, while not adjacent, have the potential for risk transmission are analyzed, and connections are established for them. Each connection edge contains rich key information: the source node and target node clearly define the starting and ending areas of the connection, clarifying the direction of risk transmission; the connection type indicates the specific nature of the relationship between nodes, whether it is a material transfer connection, an energy transfer connection, or another functional connection; the hazard attributes of the substance detail the hazardous characteristics of the substance involved in the connection, such as whether it is flammable or toxic; the probability of hazardous substance transmission is determined through data statistics, reflecting the likelihood of hazardous substances being transmitted from the source node area to the target node area. By establishing these directed connections containing multi-dimensional information, the risk transmission relationship between various functional areas in a chemical product production workshop can be accurately presented, providing a key basis for subsequent risk assessment and management.

[0044] For each connecting edge, its material safety level and hazardous substance transmission probability information are extracted. The material safety level is a pre-defined standard based on factors such as the chemical properties and hazard level of the substance itself; a higher value indicates a higher degree of danger. The hazardous substance transmission probability is determined through various methods, including statistical analysis of past accident data, simulation experiments of similar chemical environments, and expert judgment. It reflects the likelihood of a hazardous substance spreading from the starting region (source node) to the ending region (target node) of the connecting edge. Then, the material safety level and hazardous substance transmission probability are combined using a simple multiplication method: the transmission risk weight equals the material safety level multiplied by the hazardous substance transmission probability. This weight value comprehensively considers both the hazard of the substance and the probability of transmission. After obtaining the transmission risk weight for each edge, these weights are assigned to the corresponding connecting edges, so that the connecting edges not only reflect the connection relationship between regions but also intuitively reflect the magnitude of the risk carried by different connections. Finally, based on the nodes with defined attributes, the edges with connection relationships, and the weight information assigned to the edges, a weighted directed graph is constructed. In this diagram, nodes represent various functional areas of the chemical product production workshop, and directed edges between nodes clearly show the possible directions of risk transmission. The weights of the edges quantify the relative severity of risk transmission. In this way, a regional transmission topology diagram is successfully obtained, which comprehensively and intuitively depicts the risk transmission network between various functional areas within the workshop.

[0045] In one possible implementation, step S100 further includes:

[0046] Step S150: Based on the real-time process status, dynamically analyze the process reaction risk level and configure the regional risk adjustment boundary.

[0047] Step S160: Reconstruct the functional partition boundary according to the regional risk adjustment boundary, and adjust the adjacent edge weights according to the span range of the boundary reconstruction region.

[0048] Step S170: Based on the adjustment results of the node region reconstruction boundary and adjacent edge weights, perform regional time-series dynamic updates on the region transmission topology graph.

[0049] Specifically, the process is first meticulously divided, with multi-stage state recognition logic established, covering stages such as reaction start-up, stable reaction, exothermic or overpressure, and cooling. Different stages correspond to different risk levels and regional influence radius parameters, determined based on extensive historical data, process characteristic studies, and professional assessments. For example, in the reaction start-up stage, reaction conditions are initially established, reactant concentrations and reaction rates are relatively stable, resulting in a lower risk level and a smaller regional influence radius. In contrast, in the exothermic or overpressure stage, the chemical reaction is intense, heat accumulates significantly, or pressure exceeds normal ranges, easily leading to danger; therefore, the risk level is higher, and the regional influence radius is larger. Various sensors collect process parameters such as temperature, pressure, flow rate, and material concentration in real time. Based on this real-time data, the current process state can be accurately determined. Once the state is determined, the process reaction risk level is dynamically analyzed according to pre-set rules. For instance, in the exothermic reaction stage, if the temperature rises above a certain threshold and continues to rise, the risk level is determined to be elevated, and the reaction unit's warning range is automatically expanded, increasing the warning radius by 1.5-2 times. This configures the regional risk adjustment boundary and clarifies the potential range of risk spread.

[0050] Taking the chlorination reaction process in a chemical workshop as an example, its multi-stage state identification logic and parameter configuration are as follows: Reaction start-up stage (0-30 min): The process characteristics are that ethanol (initial concentration 99.5%) and chlorine (initial pressure 0.5 MPa) are injected into the reactor in a certain proportion, the temperature gradually rises to 120℃, and the reaction rate is 0.1 mol / (L·min). Based on historical data, the risk level of this stage is L (low), and the influence radius is 5 meters by default (centered on the reactor). For example, the start-up record shows that when the temperature rises to 100℃ and the pressure stabilizes at 0.3 MPa, the risk level remains L, and the influence radius does not expand. Stable reaction stage (30-180 min): The process parameters are maintained at a temperature of 180-220℃, a pressure of 1.2 MPa, a reaction rate stabilized at 0.8 mol / (L·min), and the purity of the product chloroethane is 98%. The risk level of this stage is M (medium), and the influence radius expands to 8 meters. Historical data shows that when temperature fluctuations are within ±5℃ and pressure fluctuations are within ±0.1MPa, the risk level remains at M. For example, during stable operation, with a temperature of 210℃ and a pressure of 1.15MPa, the influence radius remains unchanged. Exothermic or overpressure stage (abnormal trigger): When the sensor detects a temperature exceeding 220℃ with a heating rate >5℃ per minute, or a pressure exceeding 1.5MPa with a pressure increase rate >0.2MPa / min, a high-risk stage is determined. For example, in an abnormal event, if the temperature suddenly rises from 210℃ to 240℃ within 5 minutes (heating rate 6℃ / min) and the pressure reaches 1.8MPa, the risk level is dynamically upgraded to H (high) based on the rules, automatically expanding the reaction unit's warning radius from 8 meters to 15 meters, and including the adjacent raw material storage area (20 meters from the reactor) within the risk adjustment boundary. Cooling stage (after reaction termination): The process uses circulating cooling water (inlet temperature 25℃, flow rate 50m³ / min). 3The reactor temperature is reduced to below 80℃, the pressure returns to atmospheric pressure, and the risk level gradually decreases to L, with the radius of influence shrinking to 5 meters. During the cooling process, when the temperature drops to 70℃ and the pressure reaches 0.1MPa, the risk level returns to L, and the boundary shrinks synchronously. The distributed control system (DCS) collects process parameters in real time, including temperature (accuracy ±1℃), pressure (accuracy ±0.05MPa), flow rate (ethanol flow meter range 0-300L / h, accuracy ±1%), and material concentration (chlorine detector range 0-1000ppm, resolution 1ppm). For example, upon receiving temperature sensor data of 235℃ (exceeding the threshold of 220℃) and a heating rate of 7℃ / min, combined with pressure sensor data of 1.6MPa (exceeding the threshold of 1.5MPa) and a pressure increase rate of 0.3MPa / min, the system determines that the reactor has entered an exothermic overpressure stage, triggering a risk level escalation logic. At this point, according to preset rules, the risk adjustment boundary of the reaction unit area is expanded from an 8-meter radius centered on the original coordinates (X25-Y15) to a 15-meter radius, covering the power auxiliary facility area 10 meters to the east. Simultaneously, a risk warning is issued to the adjacent raw material storage area (20 meters from the reactor, outside the 15-meter boundary, but requiring monitoring for subsequent transmission). This dynamic configuration process is implemented through a PLC (Programmable Logic Controller), with a response time of ≤2 seconds from parameter anomalies to boundary adjustment, ensuring accurate definition of the risk diffusion range.

[0051] After determining the regional risk adjustment boundaries, the functional zoning boundaries are restructured. Taking a critical anomaly phase with rapidly increasing pressure as an example, adjacent functional units are temporarily merged into high-risk areas for unified management based on the risk adjustment boundaries. This is because, under such high-risk conditions, adjacent areas are highly susceptible to risk transmission, and including them in unified management allows for more effective risk control. Simultaneously, the weights of adjacent edges are adjusted based on the span of the restructured boundary area. A larger span means a greater potential impact from risk transmission. For example, if one of the adjacent functional areas is included in a high-risk area after boundary restructuring, and its span is large, the risk weight of this adjacent edge will be increased by more than 50% according to the rules. The weight adjustment range is precisely calculated based on specific span values, the degree of correlation between areas, and other factors, thereby quantifying the changes in risk transmission between areas and making risk assessment more accurate.

[0052] Based on the reconstructed boundaries of node regions and the adjustment results of adjacent edge weights, the regional transmission topology diagram is updated. In the topology diagram, nodes represent various functional regions, and their attributes reflect the risk status of the region, such as risk level and whether it is in a high-risk state. Connecting edges represent the relationships between regions, and their weights reflect the degree of risk transmission. Node attributes and connecting edge weights are dynamically redrawn. For example, a node that was originally in a normal state may change its color, shape, and other attributes after the process state changes to a high-risk stage, to visually display the change in its risk status; the thickness or color of connecting edges will also change according to the weight adjustment, with higher weights resulting in thicker edges or more prominent colors. Each adjustment brought about by a change in process state is synchronized to the topology diagram in real time, forming a regional transmission diagram with time-series characteristics that reflects the risk evolution process. In this way, the safety identification model and early warning decision module can obtain the latest and most accurate risk distribution information, effectively identify hazards, and make reasonable early warning decisions to ensure the safe and stable operation of chemical production.

[0053] In one possible implementation, step S300 further includes:

[0054] Step S310: Use historical accident record data and industry standards to set safety classification rules.

[0055] Step S320: Construct training datasets and validation datasets according to the safety sample characteristics, safety types, and safety level labels in the historical accident record data.

[0056] Step S330: Use the training dataset and validation dataset to train and validate the initial classifier to obtain the classifier. The classifier is used to input monitoring data features and output security type and security level.

[0057] Specifically, historical accident records are collected. This data originates from various production scenarios within the chemical industry, covering accident information from chemical enterprises of different sizes and process types during past production processes. Details include the time, location, source of the accident, and degree of harm caused. Simultaneously, industry standards are strictly adhered to. These standards, formulated by industry associations and regulatory departments, comprehensively consider the safety, standardization, and risk control requirements of chemical production. Based on this, safety is classified according to the patterns presented in the historical accident records and the guidance of industry standards. The classification process focuses on several key safety attributes. For example, hazards that are highly flammable, explosive, or toxic, and that could easily lead to large-scale explosions, fires, or severe environmental pollution if out of control during production, are classified as high-level. Hazards that are relatively stable under normal production conditions, and whose impact is limited to a small area or minor harm even in the event of anomalies, are classified as low-level. Through a comprehensive weighing of factors such as the nature of safety, potential hazard level, and probability of accident occurrence, a scientific, reasonable, and practical safety classification rule is established.

[0058] Based on the established safety grading rules, historical accident records were thoroughly mined and organized. Safety sample features were extracted from the massive dataset, encompassing multi-dimensional information such as material properties (e.g., chemical composition, ignition point, toxicity concentration thresholds), equipment operating parameters (abnormal fluctuations in temperature, pressure, and rotational speed), and process operation conditions (material ratio deviations, uncontrolled reaction times). Simultaneously, based on the actual accident situation, each sample was precisely labeled with its corresponding safety type, such as uncontrolled chemical reaction, equipment failure, or human error, and its safety level (high, medium, and low) determined according to the grading rules. Subsequently, the ratio of training to validation datasets was set at 7:3. The training dataset is primarily used for subsequent classifier training, allowing the classifier to learn the mapping relationship between different features and safety types and levels through a large amount of sample data. The validation dataset plays a crucial role in verifying the training effect. During classifier training, the validation dataset is used to test the classifier, judging its prediction accuracy and stability, ensuring that the classifier can accurately identify safety types and determine safety levels in practical applications, thereby providing reliable assurance for chemical production safety.

[0059] The Gradient Boosting Decision Tree (GBDT) algorithm was used to train and optimize the classifier. First, the safety sample features (such as abnormal fluctuations in reaction temperature, pressure exceeding limits, material toxicity concentration, equipment vibration frequency, etc.), labeled safety types (chemical reaction runaway, equipment failure, etc.), and safety levels (high, medium, low) from the training dataset were used as input. The GBDT algorithm, based on an additive model and a forward stepwise algorithm, starts with an initial simple decision tree. In each iteration, it fits the residual between the previous prediction and the true value, constructing a new decision tree by minimizing a loss function (such as a logarithmic loss function, suitable for classification problems) to correct previous prediction biases. During training, focusing on the hazardous characteristics of chemical production, the algorithm prioritizes optimizing the feature weights related to the hazard attributes of substances and fluctuations in process parameters. For example, when multiple accident samples caused by a sudden rise in reaction temperature appear in the training data, the algorithm increases the weight of the temperature feature in the decision tree split to highlight its crucial role in safety classification. After each decision tree is built, a validation dataset is used to verify its effectiveness. The validation data is input into the current model to obtain predictions for safety type and safety level. Model performance is evaluated by calculating metrics such as accuracy and confusion matrix. If the model has low accuracy in identifying high-risk levels on the validation set, or if it misclassifies different safety types, the parameters of the GBDT algorithm, such as tree depth, learning rate, and subsampling ratio, are adjusted, and the model is retrained and validated. After multiple rounds of iterative optimization, when the model's evaluation metrics on the validation dataset no longer significantly improve and tend to stabilize, the model is considered to have converged. The trained GBDT model at this point is the required classifier. In actual chemical production scenarios, real-time monitored data features such as temperature, pressure, and material concentration are input into this classifier, and the model can quickly and accurately output the corresponding safety type and safety level, providing a scientific basis for safety risk early warning and control in chemical enterprises.

[0060] In one possible implementation, step S320 further includes:

[0061] Step S321: Obtain monitoring hazard information, wherein the monitoring hazard information is different from the safety type in the historical accident record data.

[0062] Step S322: Based on the monitored hazardous information, collect hazardous characteristics and label them with safety type and safety level, and construct incremental training set and validation set.

[0063] Step S323: Use the incremental training set and validation set to incrementally learn the classifier until the validation convergence target is reached.

[0064] Specifically, diverse monitoring equipment deployed in production workshops and storage areas, such as high-sensitivity gas detectors, infrared thermal imagers, and vibration sensors, collects real-time production environment data. Simultaneously, online analytical instruments are used to analyze the composition of newly introduced chemical raw materials and the characteristics of intermediate products in the reaction process. For example, when novel nanomaterials are used as catalysts in a reaction, the unique physicochemical properties of these materials and their activity changes during the reaction process are not found in historical accident records. Similarly, due to process upgrades, the introduction of high-temperature, high-pressure continuous flow reaction processes may result in fluid dynamic anomalies, localized overheating, and other risks that differ from previous accident types. By monitoring and analyzing data from these new scenarios, new substances, and new processes, information on hazardous materials that differ from historical safety records can be accurately identified, laying the foundation for subsequent risk prevention and control.

[0065] For monitored hazardous materials, various professional detection methods are used, such as gas chromatography-mass spectrometry and X-ray diffraction, to deeply analyze their material composition and physicochemical properties. High-precision sensors are used to collect real-time process parameters such as temperature, pressure, flow rate, and material concentration during production, as well as operating status data such as equipment vibration frequency, current, and voltage. For example, for newly introduced special chemical raw materials, not only are their inherent hazardous characteristics such as ignition point, flash point, and toxicity detected, but their dynamic characteristics such as concentration changes and compatibility with other substances during the reaction process are also monitored. After collecting the hazardous characteristic data, a team of safety experts is organized to establish safety classification rules, combining industry standards and actual production experience to label the monitored hazardous materials with safety types. For example, if the special reactivity of a new raw material leads to a risk of runaway reaction, it is labeled as a new raw material runaway reaction type; if the mechanical failure risk is caused by the new structure of the equipment, it is labeled as a special equipment failure type. At the same time, based on factors such as the potential degree of hazard and the probability of accidents, the safety level of the hazardous material is divided into high, medium, and low levels. Finally, the labeled sample data are divided into the incremental training set and the validation set in a 7:3 ratio. The incremental training set is used to allow the initial classifier to learn the features of new types of safety, while the validation set is used to test the prediction accuracy of the initial classifier after learning new knowledge, ensuring that the model can effectively identify newly emerging risks. The classifier is then obtained after training is completed.

[0066] Gradient Boosting Decision Tree (GBDT) algorithm is used for incremental learning of the initial classifier. First, the incremental training set is input into the existing GBDT classifier. GBDT is an iterative decision tree ensemble algorithm based on the forward stepwise algorithm, where each iteration generates a new decision tree to fit the residuals of the previous model. For each sample in the incremental training set, it includes previously collected hazard features, labeled safety type, and safety level. In each iteration of building a new decision tree, GBDT randomly selects a subset of these features, using metrics such as the Gini index or mean squared error to determine the optimal split point, dividing the samples into different child nodes and gradually constructing the decision tree structure. After each training iteration, a validation set is used to evaluate the performance of the current model. For samples in the validation set, input into the current GBDT model, the model makes predictions based on its internal decision tree structure, outputting the corresponding safety type and safety level. Then, the error between the prediction result and the true label is calculated; common evaluation metrics include accuracy, recall, and F1 score. If the evaluation results do not meet the preset validation convergence target—for example, if the accuracy is below a certain threshold or the F1 score does not improve significantly—some hyperparameters of GBDT will be adjusted, such as the learning rate, the maximum tree depth, and the sampling ratio of each tree. After adjustment, a new round of training and learning is performed using the incremental training set to generate a new decision tree and add it to the model. This process is repeated continuously to optimize and improve the model. Each round of training allows the model to better learn the features and classification rules of new security types. As training progresses, the model's performance on the validation set will gradually stabilize. When the evaluation metrics no longer show significant improvement and the preset validation convergence target is reached, the incremental learning process ends. The resulting GBDT model is the updated classifier, which can more accurately identify newly emerging security types and security levels.

[0067] In one possible implementation, step S400 further includes:

[0068] Step S410: Locate the hazardous area based on the location of the monitor deployment and the effective monitoring range.

[0069] Step S420: Project the security identification result onto the regional transmission topology diagram according to the location, and predict the risk transmission of adjacent regions based on the regional transmission path and transmission weight relationship to obtain the regional transmission risk.

[0070] Specifically, based on a pre-planned deployment of monitors and the effective coverage area of ​​each monitor, hazardous areas are precisely located. In complex environments such as chemical production workshops, various monitors, such as gas sensors, temperature sensors, and pressure sensors, are installed according to the process flow, equipment layout, and distribution of potentially hazardous substances. These monitors collect data in real time. Once a monitor detects abnormal data, such as excessive concentrations of harmful gases or abnormal increases in temperature or pressure, the specific area where a potential hazard may exist is immediately determined by combining the monitor's location and effective monitoring range. For example, if a gas monitor in a certain area detects a toxic gas leak, the approximate location of the leak source can be quickly pinpointed based on the monitor's location and its effective monitoring radius.

[0071] A graph propagation algorithm combined with Bayesian network concepts is employed to predict the transmission of risk to adjacent regions. First, the security identification results are accurately mapped to a regional transmission topology graph based on their location information. This graph is represented by nodes representing functional regions, edges representing connections between regions, and edge weights reflecting the probability and degree of risk transmission. Based on the graph propagation algorithm, risk information is propagated starting from nodes identified as dangerous. Initially, the risk value of dangerous nodes is set to 100% (which can be quantified according to actual conditions), while the risk value of other nodes is 0. Then, iterative propagation is performed based on the regional transmission path and transmission weights. In each iteration, each node distributes its own risk value to its neighboring nodes according to the weight ratio of the edges connected to it. For example, if node A is connected to nodes B and C, the weight of edge AB is 0.3, the weight of edge AC is 0.7, and the risk value of node A is RA, then in this iteration, node B will receive a risk increment of RA × 0.3 from node A, and node C will receive a risk increment of RA × 0.7. Simultaneously, the concept of Bayesian networks is introduced to consider the conditional probability of risk during the transmission process. For example, certain areas may have reduced probability of hazard occurrence after risk propagation due to protective measures or specific processes. A pre-defined conditional probability table is used to adjust the risk values ​​during propagation. After each iteration, the risk value of each node is updated until the risk values ​​of all nodes no longer change significantly, reaching a convergence state. Ultimately, the risk value corresponding to each node represents the regional transmission risk of that adjacent area. In this way, the potential risk level of each adjacent area can be predicted relatively accurately, providing a scientific basis for the safety management of chemical production.

[0072] In one possible implementation, step S420 further includes:

[0073] Step S421: Based on the location of the monitors and the effective monitoring range, perform cross-thermal distribution analysis of the monitoring data to determine the diffusion distribution characteristics of the risk and identify the risk center point.

[0074] Step S422: Using the risk center point as the initial node, establish a set of directed propagation paths starting from the initial node in the current segmented region propagation topology graph.

[0075] Step S423: Based on the risk propagation attenuation factor and the edge weights of the regional propagation topology graph, calculate the propagation risk value of the directed propagation path to obtain the regional propagation risk.

[0076] Specifically, based on the specific deployment locations of the monitors within the factory area and their respective effective monitoring ranges, cross-thermal distribution analysis is conducted on the collected multi-source monitoring data. Data such as temperature, pressure, and gas concentration, collected in real-time by the sensor network, are integrated using data fusion technology. Subsequently, based on a Geographic Information System (GIS), the monitoring area is divided into regular grids, and the comprehensive values ​​of various monitoring indicators within each grid are calculated. Kernel density estimation is used to process these values, generating a cross-thermal distribution map, where color intensity visually reflects the degree of hazard. The distribution pattern and gradient changes of colors in the thermal map are analyzed to determine the diffusion direction, speed, and scope of the risk. Finally, clustering analysis algorithms are used to identify the area with the most concentrated and highest values ​​in the map; the center of this area is the risk center point, providing precise location for subsequent risk prevention and emergency response.

[0077] After identifying the risk center point, this center point is used as the key initial node for analysis within the constructed and currently segmented regional transmission topology graph. The regional transmission topology graph uses nodes to represent different functional areas in chemical production, and the connecting edges between nodes represent the risk transmission relationships between areas. The edge weights reflect the probability and intensity of risk transmission. Starting from the node corresponding to the risk center point, a search is performed along the edges based on the connection relationships between nodes in the graph. Using depth-first search (DFS) or breadth-first search (BFS) algorithms in graph theory, all reachable nodes are traversed, thus constructing a set of directed propagation paths originating from the initial node. These directed propagation paths clearly demonstrate which adjacent areas and further regions the risk may propagate from the center point, providing a structured data foundation for subsequent quantitative assessment of the transmission risks faced by each area. This helps to intuitively grasp the potential paths of risk spread, enabling the early development of targeted prevention and control strategies.

[0078] First, it is clarified that the risk propagation attenuation factor reflects the degree to which the risk intensity weakens during the propagation process due to factors such as distance, obstacles, and protective measures. The edge weights in the regional transmission topology diagram reflect the probability and difficulty of risk transmission between two regions. For the derived set of directed propagation paths, starting from the risk center point as the initial node, its risk value is set to the highest preset value. Along each path, when the next node receives risk, the current node's risk value is first multiplied by the weight of the edge connecting this node and the next node to obtain the preliminary transmission risk value. Then, it is multiplied by the propagation attenuation factor to correct it, calculating the actual transmission risk value of the next node. In this way, the risk value is calculated for all nodes on the path. When the node risk values ​​of all paths have been calculated, if multiple paths point to the same node, the risk value of that node is aggregated using a weighted summation method. Finally, by combining the final risk values ​​of all nodes, the transmission risk of the entire region is obtained, providing a quantitative basis for chemical production safety management.

[0079] Table 1: An example of calculating the transmission risk value of a directed propagation path based on the risk propagation attenuation factor and the edge weights of the regional transmission topology graph:

[0080] Table 1

[0081]

[0082] Example 2, based on the same inventive concept as the safety monitoring and management method for chemical product production workshops in the foregoing examples, such as... Figure 2 As shown, this application provides a safety monitoring and management system for chemical product production workshops. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0083] The topology diagram creation module 10 is used to create a regional transmission topology diagram for the chemical product production workshop according to the regional functions.

[0084] The multi-source detection data acquisition module 20 is used to deploy various types of monitors at functional nodes such as reaction units, raw material storage areas, high-temperature and high-pressure equipment areas, and power auxiliary facilities to collect multi-source detection data.

[0085] The security identification result localization module 30 is used to train a classifier based on security level classification rules, perform security identification classification through the classifier based on the multi-source detection data, and locate the security identification result.

[0086] The early warning level generation module 40 is used to perform regional point identification and regional transmission risk prediction based on the regional transmission topology map and the safety identification results, generate an early warning level, and perform graded early warning and emergency management control of the product production process based on the early warning level.

[0087] Furthermore, the system is also used to implement the following functions:

[0088] The various types of monitors include: combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors. The monitor deployment nodes in the functional areas include one or more of the combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors.

[0089] Furthermore, the system is also used to implement the following functions:

[0090] The functional areas are analyzed based on their characteristics, including process flow, equipment correlation, spatial location, material properties, and personnel and energy flow. Each functional area is defined as a node, with node attributes including node number, functional type, and basic information on material or energy risk level. Directed connections are established between nodes based on process flow, equipment correlation, and spatial location. Each edge includes source and target nodes, connection type, material hazard properties, and hazardous material transmission probability. Based on the material safety level and hazardous material transmission probability in the connection edges, the transmission risk weight of each edge is calculated and assigned a weight. A weighted directed graph is then constructed based on the nodes, edges, and edge weights to obtain the transmission topology diagram of the region.

[0091] Furthermore, the system is also used to implement the following functions:

[0092] Based on the real-time process status, the process reaction risk level is dynamically analyzed, and regional risk adjustment boundaries are configured. Functional partition boundaries are reconstructed based on the regional risk adjustment boundaries, and adjacent edge weights are adjusted based on the span of the boundary reconstruction region. Based on the adjustment results of the node region reconstruction boundaries and adjacent edge weights, the regional transmission topology diagram is dynamically updated in a time sequence.

[0093] Furthermore, the system is also used to implement the following functions:

[0094] Using historical accident record data and industry standards, safety classification rules are set; according to the safety sample characteristics, safety type, and safety level labels in the historical accident record data, training datasets and validation datasets are constructed; the initial classifier is trained and validated to converge using the training datasets and validation datasets to obtain the classifier, which is used to input monitoring data features and output safety type and safety level.

[0095] Furthermore, the system is also used to implement the following functions:

[0096] Obtain information on monitored hazardous materials, wherein the safety type of the monitored hazardous materials is different from that in historical accident record data; collect hazard features based on the monitored hazardous materials information and label them with safety type and safety level, and construct incremental training set and validation set; use the incremental training set and validation set to incrementally learn the classifier until the validation convergence target is reached.

[0097] Furthermore, the system is also used to implement the following functions:

[0098] Based on the location of the monitors and their effective monitoring range, dangerous areas are located. The safety identification results are then projected onto the regional transmission topology diagram according to the location. Based on the regional transmission path and transmission weight relationship, the risk transmission of adjacent areas is predicted to obtain the regional transmission risk.

[0099] Furthermore, the system is also used to implement the following functions:

[0100] Based on the deployment location and effective monitoring range of the monitors, cross-thermal distribution analysis of the monitoring data is performed to determine the diffusion distribution characteristics of the risk and identify the risk center point. Using the risk center point as the initial node, a set of directed propagation paths starting from the initial node is established in the regional transmission topology graph of the current segmentation state. Based on the risk propagation attenuation factor and combined with the edge weights of the regional transmission topology graph, the transmission risk value of the directed propagation paths is calculated to obtain the regional transmission risk.

[0101] Example 3: Based on the same inventive concept as the safety monitoring and management method for chemical product production workshops in the preceding examples, this example provides a computer-readable storage medium that can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the safety monitoring and management method for chemical product production workshops in this application. The processor executes the software programs, instructions, and modules stored in the memory to perform various functional applications and data processing of the computer device, thereby realizing the aforementioned safety monitoring and management method for chemical product production workshops.

[0102] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0103] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0104] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A safety monitoring management method for a chemical product production plant, characterized by, The method comprises the following steps: establishing a regional conduction topology graph according to the functional division of a chemical product production workshop; deploying multiple types of monitors at the nodes of the functional areas of the reaction unit, raw material storage area, high-temperature and high-pressure equipment area, and power auxiliary facility to collect multi-source detection data; training a classifier based on safety level division rules, performing safety identification and classification through the classifier according to the multi-source detection data, and locating the safety identification result; performing regional fixed-point identification and regional conduction risk prediction according to the safety identification result positioning, generating an early warning level, and performing hierarchical early warning and product production process emergency management control according to the early warning level; the multiple types of monitors include combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors, wherein the monitor node of the functional area includes one or more of the combustible gas detectors, toxic gas detectors, temperature sensors, pressure sensors, vibration sensors, and video monitors; establishing a regional conduction topology graph, comprising: analyzing the characteristic relationship of each functional area according to the process flow, equipment correlation, spatial position relationship, material attribute, personnel, and energy flow; defining each functional area as a node, and the node attribute includes node number, function type, and material or energy risk level basic information; establishing a directed connection edge between nodes according to the process flow, equipment correlation, and spatial position relationship, and each edge includes a source node and a target node, a connection type, a material hazard attribute, and a hazardous material conduction probability; calculating the conduction risk weight of each edge according to the material safety level and the hazardous material conduction probability in the connection edge, assigning the edge weight, and establishing a weighted directed graph based on the nodes and edges and the edge weight to obtain the regional conduction topology graph; establishing a regional conduction topology graph, further comprising: dynamically analyzing the process reaction risk level according to the real-time process state, and configuring the regional risk adjustment boundary; reconfiguring the functional partition boundary according to the regional risk adjustment boundary, and adjusting the adjacent edge weight according to the boundary reconfiguration area span range; performing regional time sequencing dynamic update on the regional conduction topology graph according to the adjustment results of the node area reconfiguration boundary and the adjacent edge weight.

2. The safety monitoring management method of a chemical plant according to claim 1, characterized by, Training a classifier based on safety level division rules, comprising: setting safety classification rules using historical accident record data and industry standards; constructing a training data set and a validation data set according to the safety sample features, safety types, and safety levels in the historical accident record data; training and validating the initial classifier using the training data set and the validation data set to obtain the classifier, which is used to input the monitoring data features and output the safety type and safety level.

3. The safety monitoring management method of a chemical plant according to claim 2, characterized by, Constructing a training data set and a validation data set, comprising: obtaining monitoring hazard information, which is different from the safety type in the historical accident record data; collecting hazard features based on the monitoring hazard information and labeling the safety type and safety level to construct an incremental training set and a validation set; Incrementally train the classifier using the incremental training set and the validation set until a validation convergence goal is reached.

4. The safety monitoring management method of a chemical industrial products manufacturing plant according to Claim 1, characterized by, Generate an early warning level, and perform hierarchical early warning and emergency management control of the product production process based on the early warning level, including: Positioning a safety risk area based on the monitoring device layout position and the monitoring effective range; Projecting the safety identification result into the area conduction topology structure diagram according to the positioned safety risk area, and performing adjacent area risk conduction prediction based on the area conduction path and the conduction weight relationship to obtain an area conduction risk; Jointly determining the early warning level according to the point safety identification risk and the area conduction risk of the safety risk area, wherein the early warning level is the cumulative result of the point early warning level and the conduction risk area; Based on the point safety identification risk and the area conduction risk, performing dynamic product timing management control and diffusion timing response control according to the risk identification result to ensure that each functional area meets the preset safety target.

5. The safety monitoring management method of a chemical industrial products manufacturing plant according to claim 4, characterized by, To obtain the area conduction risk, further comprising: According to the monitoring device layout position and the monitoring effective range, performing cross-thermal distribution analysis of the monitoring data to determine the diffusion distribution characteristics of the risk and identify the risk center point; Taking the risk center point as an initial node, establishing a set of directed propagation paths starting from the initial node in the area conduction topology structure diagram in the current segmentation state; According to the risk propagation attenuation factor, combining the edge weight of the area conduction topology structure diagram, and calculating the conduction risk value of the directed propagation path to obtain the area conduction risk.

6. A safety monitoring management system for a chemical production plant, characterized in that, The system is used to implement the safety monitoring management method of the chemical product production plant according to any one of claims 1-5, and the system comprises: A topology structure diagram establishing module for dividing the chemical product production plant according to the area function and establishing an area conduction topology structure diagram; A multi-source detection data acquisition module for laying multiple types of monitoring devices at the reaction unit, raw material storage area, high temperature and high pressure equipment area, and power auxiliary facility functional area nodes, and acquiring multi-source detection data; A safety identification result positioning module for training a classifier based on a safety level division rule, performing safety identification classification based on the multi-source detection data through the classifier, and positioning the safety identification result; An early warning level generation module for generating an early warning level based on the area conduction topology structure diagram, combining the safety identification result positioning, performing area point identification and area conduction risk prediction, and performing hierarchical early warning and emergency management control of the product production process based on the early warning level.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that The program is executed by the processor to implement the safety monitoring management method of the chemical product production plant according to any one of claims 1-5.

Images