Intelligent management and control system and method for safety risk of chemical industry park

Abstract

The invention discloses a chemical industry park safety risk intelligent management and control system and method, belongs to the technical field of risk management and control, and solves the problems of incomplete tank field risk perception, data transmission delay and insufficient dynamic path risk avoidance. The system comprises a sensing device; the 5G communication module and the industrial Ethernet form a dual-mode transmission network to upload data to a cloud platform; the cloud platform uses a time sequence database to establish an equipment code index, and a rule engine executes leakage judgment of methane for continuous 3-5 cycles gt; 25% LEL and a pressure drop rate gt; early warning is triggered at the speed of 0.5 MPa / min, a risk analysis module calls three-dimensional geographic information data to generate an evacuation path with the radius of 500-2000 m through a Dijkstra algorithm, an upwind area of the real-time wind direction is avoided, and the fire station is preferentially connected; and the application terminal receives an early warning instruction containing the leakage equipment code and the dynamic path. The system is used for rapid early warning and evacuation planning of chemical safety accidents.

Description

Technical Field

[0001] The present invention relates to the technical field of risk control in chemical industrial parks, and particularly to an intelligent risk control system and method for chemical industrial park safety risks. Background Art

[0002] There are problems of lagging response and low evacuation efficiency in the safety risk monitoring of major hazard source tank farms in chemical industrial parks, which are mainly caused by multi-dimensional technical defects. At the risk perception level, existing systems usually only use a single type of sensor for monitoring. For example, only relying on gas concentration detection or pressure detection, it is difficult to capture multi-parameter coupling risks. Since the early stage of gas leakage often shows the composite characteristics of sudden pressure drop and slow concentration rise, a single sensor cannot establish the association rules between parameters, resulting in a high false negative rate of early leakage warning. At the same time, the deployment positions of traditional sensors are mostly fixed outside the tank body, and they respond slowly to the local concentration gradient changes caused by minor leaks, delaying the best disposal opportunity.

[0003] In the data transmission link, existing systems generally rely on a single network mode for communication. When using a wireless network, the electromagnetic shielding effect generated by the dense metal tank bodies in the chemical industrial park is likely to cause signal attenuation, and packet loss frequently occurs during high-concurrency data transmission. The change trend of key parameters (such as the instantaneous pressure drop rate) is difficult to be uploaded completely; while when using a wired industrial bus, the network scalability is poor. Adding new monitoring points requires re-wiring, and the bandwidth allocation is rigid, unable to adapt to the traffic peak under sudden risk events. These defects result in time stamp misalignment or key frame loss in the data received by the cloud platform, directly affecting the timeliness of subsequent risk analysis.

[0004] In the risk disposal stage, there are obvious limitations in existing path planning methods. On the one hand, the evacuation path is generated relying on static electronic map data and does not integrate meteorological monitoring information in real time. Since the diffusion range of chemical gases is directly affected by wind direction and wind speed, the static path may guide people into the gas diffusion area, causing secondary injuries. On the other hand, the path planning is not dynamically associated with emergency resources, for example, it does not give priority to connecting to the nearest available fire station or medical point. The path for rescue personnel to reach the leakage point is circuitous, delaying the initial disposal efficiency. In addition, the evacuation instructions issued by traditional systems lack detailed spatial information, only providing text warnings or simple schematic diagrams, and on-site personnel cannot quickly locate the geometric relationship between their own positions and the optimal path.

[0005] These problems stem from three technical bottlenecks: First, the real-time fusion of multi-source heterogeneous sensor data requires addressing the parallel processing challenges of high-throughput time-series data. In particular, parameters such as gas concentration, pressure, and temperature have different sampling frequencies and dimensions. Establishing cross-parameter association rules requires overcoming the complexity of time series alignment and feature extraction. Second, dynamic path planning requires overcoming computational barriers to efficient coupling of meteorological data and geographic information systems. The impact of wind direction changes on paths requires a response within seconds, but traditional spatial analysis algorithms are computationally expensive. Finally, the simultaneous issuance of warning instructions and evacuation routes requires balancing transmission reliability and real-time performance. High-precision vector maps are data-intensive, and achieving low-latency transmission within limited bandwidth presents technical challenges. These difficulties have long constrained improvements in the safety management and control effectiveness of chemical parks. Summary of the Invention

[0006] This invention provides an intelligent safety risk management and control system and method for chemical parks. These systems aim to address issues such as incomplete multi-parameter risk perception in tank areas with major hazardous sources in chemical parks, insufficient real-time data transmission, and a failure to integrate real-time weather and emergency resources into dynamic evacuation routes. Existing systems, due to their single sensor, rigid network architecture, and static path planning, result in high leak warning misses, evacuation routes crossing hazardous areas, and delayed rescue responses. The system addresses the issue of evacuation routes being rendered inoperable by obstacles. Traditional systems rely on preset maps, making route replanning triggered by new obstacles (such as collapsed facilities) inefficient and unable to guarantee route availability. The system addresses the congestion in issuing warning commands when large numbers of terminals access the system concurrently. Existing broadcast-based distribution systems do not prioritize risk proximity, and terminals in high-risk areas may be delayed in receiving critical path updates due to network congestion. The system addresses the slow response of manual response processes. Traditional methods rely on manual judgment of leak characteristics and manual route drawing, resulting in a timeframe of more than minutes from data collection to command issuance, failing to meet the second-by-second emergency response requirements. The system also addresses the issue of dynamic meteorological changes weakening route safety. When wind speed and direction change, fixed-angle avoidance zones fail to match the gas diffusion range, causing the route to still pass through contaminated areas. Solve the problem of low efficiency of coordinated evacuation at multiple leakage points. Independent path planning does not take into account the aggregation effect of leakage points, and scattered paths may cross and conflict, increasing the overall evacuation time. Solve the problem of increased risk of congestion in main evacuation channels. There is a lack of automatic diversion mechanism when the population density exceeds the limit. The traditional system only provides a single path, which easily forms a bottleneck. Solve the problem of deviation in path execution of on-site personnel. In complex environments, people are prone to lose their way. Static path maps cannot dynamically correct the trajectory of action, and there is no real-time intervention after deviation. Solve the problem of heavy protective equipment hindering people from perceiving alarms. The sound insulation of chemical protective clothing causes voice prompts to fail, and the vibration feedback is not adapted to the weight difference of equipment, so key steering instructions may be missed. Solve the problem of decreased recognition rate of alarm signals during intense exercise. The bumps of running reduce the perceptibility of conventional vibration / light signals, and motion interference is not dynamically compensated.

[0007] To achieve these objects and other advantages in accordance with the present invention, an intelligent management and control system for chemical industrial park safety risks is provided, including perception layer devices, a data transmission network, a cloud platform data processing center, and application terminals deployed in the chemical industrial park; The perception layer devices include gas concentration sensors, temperature sensors, pressure sensors, and flame detectors provided in major hazard source tank farms; The data transmission network adopts a dual-mode transmission architecture of 5G communication modules and industrial Ethernet, and the perception layer devices package the collected data and transmit it to the cloud platform data processing center; The cloud platform data processing center includes a time series database, a rule engine, and a risk analysis module. The time series database stores the collected data of the perception layer devices in timestamp format and establishes an equipment coding index. The rule engine pre-sets a leakage determination rule: when the methane concentration data associated with the same equipment coding exceeds 25% LEL for 3 to 5 consecutive sampling periods, and the pressure data decline rate is greater than 0.5 MPa / min, a leakage warning is triggered; after receiving the warning signal output by the rule engine, the risk analysis module calls the equipment coordinate data, real-time meteorological data, and emergency resource distribution data in the three-dimensional geographic information system, and calculates the evacuation path within a range of 500 meters to 2000 meters centered on the leakage point through the Dijkstra algorithm. The evacuation path avoids the upwind area of the real-time wind direction and preferentially connects to the fire station coordinate point; The application terminal receives the warning instructions and evacuation path vector map issued by the cloud platform data processing center. The warning instructions include the leakage equipment coding, warning level, and gas diffusion simulation polygon. The evacuation path vector map includes a dynamically updated set of path node coordinates and the estimated passing time of each node.

[0008] Preferably, in the intelligent management and control system for chemical industrial park safety risks of the present invention, the risk analysis module further includes a lidar scanning unit, which continuously acquires real-time point cloud data of the path node area after generating the evacuation path; the scanning frequency of the lidar scanning unit is 5 Hz - 10 Hz, and the scanning angle range covers 30 degrees to 60 degrees on both sides of the evacuation path center line; The risk analysis module performs differential comparison between the real-time point cloud data and the base map elevation data of the three-dimensional geographic information system. When it is identified that the absolute difference between the surface point cloud elevation value of the newly added obstacle and the base map elevation data is greater than 50 cm, an evacuation path replanning is triggered; the A-Star algorithm is used in the replanning process to recalculate the obstacle avoidance path, and the updated evacuation path vector map is synchronized to the application terminal within 3 seconds to 8 seconds.

[0009] Preferably, in the intelligent control system for chemical industrial park safety risks of the present invention, the cloud platform data processing center further includes an instruction distribution module connected to both the risk analysis module and the application terminal. The instruction distribution module includes: A terminal positioning unit that obtains physical location data through the GPS positioning module built into the application terminal, with a positioning accuracy error of less than 3 meters; A distance grading unit that calculates the straight-line distance between each application terminal and the leakage point and divides it into 3 priority levels. Among them, the distance less than or equal to 500 meters from the leakage point is the first priority, the distance greater than 500 meters and less than or equal to 1000 meters from the leakage point is the second priority, and the distance greater than 1000 meters and less than or equal to 2000 meters from the leakage point is the third priority. When the distance is greater than 2000 meters, the early warning is automatically lifted; A transmission control unit that, when the evacuation path replanning is triggered, directly transmits the evacuation path vector map to the first-priority application terminal, which is sent through the 5G communication module using the UDP protocol. The data packets transmitted to the second- and third-priority application terminals are stored in the Redis cache queue and sent through the industrial Ethernet after a delay of 2 to 5 seconds. The resource allocation ratio of the transmission control unit is dynamically adjusted according to the personnel evacuation priority strategy in the chemical industrial park emergency plan. Among them, the resource allocation ratio for the first-priority application terminal accounts for 60% - 80% of the total bandwidth, and the remaining bandwidth is allocated according to the ratio of the number of second-priority application terminals: the number of third-priority application terminals = 3:1.

[0010] The present invention also provides an intelligent control method for chemical industrial park safety risks, including the following steps: Step 1, real-time collection of methane concentration data through gas concentration sensors deployed in major hazard source tank farms; Step 2, collection of tank farm temperature data through temperature sensors; Step 3, collection of tank farm pressure data through pressure sensors; Step 4, collection of flame radiation data through flame detectors; Step 5, sending the methane concentration data, temperature data, pressure data, and flame radiation data collected in Steps 1 to 4 to the cloud platform data processing center through a dual-mode transmission architecture of 5G communication module and industrial Ethernet; Step 6, storing the received data in the time series database of the cloud platform data processing center and establishing a device code index in timestamp format; Step 7, performing leakage determination through a rule engine: generating a leakage warning signal when the methane concentration data associated with the same device code exceeds 25% LEL in 3 to 5 consecutive sampling periods and the pressure data decline rate is greater than 0.5 MPa / min; Step eight, in response to the leakage warning signal, call the device coordinate data, real-time meteorological data, and emergency resource distribution data stored in the three-dimensional geographic information system; Step nine, set an evacuation route avoidance area based on the real-time wind direction data. The evacuation route area covers a 30-degree to 60-degree fan-shaped area upwind of the leakage point and a range of 200 meters to 800 meters from the leakage point; Step ten, calculate the evacuation routes within a range of 500 meters to 2000 meters centered on the leakage point through the Dijkstra algorithm. The evacuation routes bypass the evacuation route avoidance area set in step nine and are preferentially connected to the nearest fire station coordinate points; Step eleven, encapsulate the leakage device code, warning level, and gas diffusion simulation polygon into a warning instruction; Step twelve, encapsulate the evacuation route node coordinate set, the estimated passing time of each node, and the real-time wind direction vector into an evacuation route vector map; Step thirteen, send the warning instruction and the evacuation route vector map to the application terminal through the 5G communication module.

[0011] Preferably, in the intelligent control method for chemical industrial park safety risks of the present invention, step eight further includes: extracting wind speed data and wind direction angle data from the real-time meteorological data. The wind speed detection range is 1 m / s to 30 m / s, and the wind direction angle detection range is 0 degrees to 360 degrees; taking 15 seconds to 40 seconds as the interpolation period, performing cubic spline interpolation calculation on the wind speed data and the wind direction angle data to generate a wind field model with a continuous time series; associating the output of the wind field model with the leakage point device code and storing it as dynamic meteorological vector data; The path avoidance area set in step nine is updated based on the dynamic meteorological vector data. When the wind speed data is 8 m / s to 15 m / s, the fan-shaped area angle expands to a range of 45 degrees to 90 degrees; The calculation frequency of the evacuation routes in step ten is increased to be executed once every 10 seconds to 30 seconds.

[0012] Preferably, in the intelligent control method for chemical industrial park safety risks of the present invention, step eight further includes: retrieving the device codes associated with all current leakage warning signals, extracting the coordinate data of each leakage point and the dynamic meteorological vector data; establishing a buffer area with a radius of 30 meters to 60 meters centered on the leakage point coordinates, and using the spatial overlay analysis algorithm to detect the overlapping part of the buffer area; when the overlapping area ratio exceeds 40% to 70%, merging the dynamic meteorological vector data and generating a combined path avoidance area; The set range of the path avoidance area in step nine is extended to: maintaining a 30-degree to 60-degree fan-shaped area for a single leakage point; adjusting to a 50-degree to 120-degree composite fan-shaped area for the combined path avoidance area, and taking the maximum range of 200 meters to 800 meters among the leakage points; In Step Ten, the evacuation route calculation adopts a hierarchical planning strategy: first, plan the main evacuation channels to connect the safety exit coordinate points within a radius of 500 meters to 2000 meters; then generate branch paths based on the main evacuation channels to each equipment coding area.

[0013] Preferably, in the intelligent control method for safety risks in chemical industrial parks of the present invention, the branch path generation stage in Step Ten includes the following steps: S1. Obtain the real-time video monitoring data along the main evacuation channels, and identify the personnel density and moving speed in the main evacuation channels through the YOLOv5 target detection algorithm; S2. When the personnel density is 1 person per square meter - 3 people per square meter and the moving speed is 0.8 meters per second - 1.2 meters per second, mark the section of the main evacuation channel as a congestion node; S3. Retrieve the alternative branch paths within a radius of 50 meters - 100 meters from the congestion node, and calculate the detour distance increment and slope data of the alternative branch paths; S4. Select the alternative branch path with a detour distance increment of 120 meters - 200 meters and a slope of 8 degrees - 15 degrees as the diversion path; S5. Topologically associate the coordinates of the diversion path nodes with the main evacuation channel nodes to generate a dynamic evacuation road network; The topological relationship data of the dynamic evacuation road network is encapsulated in the evacuation path vector map in Step Twelve.

[0014] Preferably, in the intelligent control method for safety risks in chemical industrial parks of the present invention, Step Twelve further includes: parsing the topological relationship data of the dynamic evacuation road network, and extracting the connection node coordinates of the diversion path and the main evacuation channel; deploying LoRa wireless beacons within a range of 20 meters - 50 meters from the connection node, with a beacon broadcast frequency of 1 time per second - 3 times per second, and the broadcast content includes node coding and path direction angle; after the application terminal receives the LoRa beacon data, compare it with the dynamic evacuation road network vector map stored locally: if the distance between the current position of the application terminal and the connection node is 15 meters - 30 meters and the direction angle deviation is 20 degrees - 40 degrees, trigger a voice turning prompt, and if the moving trajectory of the application terminal deviates from the dynamic evacuation road network planned path for 10 seconds - 20 seconds continuously, start a vibration alarm and retransmit the dynamic evacuation road network vector map generated and updated in real time by the cloud platform data processing center to the application terminal; the voice prompt content is generated by using a TTS engine to convert the path instruction, including the turning angle value and a countdown within the range of 15 seconds - 30 seconds.

[0015] Preferably, in the intelligent control method for the safety risks of the chemical industrial park of the present invention, step twelve further includes: when the voice steering prompt is triggered, the gyroscope and accelerometer of the application terminal are synchronously activated, and the sampling frequency is 50Hz - 100Hz; the angular velocity value of the steering action of the application terminal holder is calculated in real time, and the effective steering threshold is set to 15 degrees / second - 30 degrees / second; If no effective steering action is detected within 3 seconds - 8 seconds after the countdown starts, the remaining countdown is extended by 5 seconds - 10 seconds; if the cumulative extension times reach 2 times - 4 times, the steering voice prompt is turned off and the vibration alarm is started; The vibration alarm mode is adjusted according to the type of protective equipment bound to the application terminal: when the protective equipment is a heavy chemical protective suit, 3 times - 5 times of long vibration lasting for 1 second - 3 seconds is adopted; when the protective equipment is a light protective suit, 8 times - 12 times of short vibration lasting for 0.3 seconds - 0.6 seconds is adopted; the vibration intensity linearly increases with the environmental noise decibel value, and the reference intensity is 1.5G - 2.5G, and the intensity increases by 0.2G - 0.5G for every 10 decibels - 15 decibels increase in noise.

[0016] Preferably, in the intelligent control method for the safety risks of the chemical industrial park of the present invention, step twelve further includes: the three-axis acceleration data is collected in real time through the accelerometer of the application terminal, and the sampling frequency is maintained at 50Hz - 100Hz; when the standard deviation of the Z-axis acceleration exceeds 0.5G - 8G within 3 seconds - 5 seconds continuously, it is determined that the application terminal is in a violent motion state; In the violent motion state, the vibration alarm and the LED indication module are synchronously activated: when the protective equipment is a heavy chemical protective suit, the red light pulse frequency is increased to 3Hz - 5Hz, and the vibration intensity is increased by 0.5G - 1.0G; when the protective equipment is a light protective suit, the green light pulse frequency is increased to 5Hz - 8Hz, and the vibration mode is switched to intermittent strong vibration, and the working cycle is to vibrate once every 0.2 seconds to 0.5 seconds, and the interval between each vibration is 0.1 seconds to 0.3 seconds; The vibration intensity compensation value is dynamically adjusted according to the motion intensity, including: the reference intensity is the set value in the protective equipment adaptation stage; for every 0.2G - 0.4G increase in the standard deviation of the Z-axis acceleration, the vibration intensity is increased by 0.1G - 0.3G; The brightness of the light pulse increases with the smoke concentration. For every 100mg / m³ - 200mg / m³ increase in the smoke concentration, the LED output lumen value is increased by 30 lumens - 80 lumens.

[0017] The present invention has at least the following beneficial effects: 1. Establish a leakage composite determination rule through multi-parameter sensor coupling (gas, temperature, pressure, flame), effectively reducing the false alarm rate of early warning; the dual-mode network architecture ensures that key data is uploaded within 5 ms, and the integrity rate of the pressure change rate of time-sensitive parameters reaches 99%; dynamic path avoidance bypasses the upwind area and preferentially connects to the fire station, increasing the compliance rate of the safe distance of the evacuation path to 95% and accelerating the rescue response speed by 40%.

[0018] 2. Laser radar point cloud differential comparison realizes the identification of new obstacles with an accuracy of 50 cm, and compresses the path replanning response time within 8 seconds; the obstacle avoidance path calculation efficiency of the A-Star algorithm is 3 times higher than that of traditional methods, ensuring the real-time availability of the updated path in complex scenarios, and reducing the path failure risk by 70%.

[0019] 3. Hierarchical management of terminal priorities reduces the path update delay of the first-priority terminal within 500 meters to less than 100 ms; Redis queue buffering combined with the dynamic bandwidth allocation strategy, with the first priority accounting for 60%-80%, avoids network congestion, and the instruction arrival rate of terminals in high-risk areas reaches 100%; network broadband resources are allocated according to a 3:1 ratio to optimize the transmission efficiency of the second and third priorities, and the overall system concurrency carrying capacity is doubled.

[0020] 4. The standardized data acquisition process (Steps 1 to 5) ensures the alignment of multi-source data timestamps, reducing the leakage determination delay from the minute level to within 10 seconds; the automation of path dynamic calculation (Step 10) and instruction encapsulation (Steps 11 to 12) increases the evacuation plan generation efficiency by 5 times; the control delay of the 5G module is <800 ms, which is 90% faster than traditional manual handling.

[0021] 5. The wind field model generates continuous meteorological vectors through cubic spline interpolation, and the avoidance area angle is dynamically adjusted with the wind speed, expanding from 8 m / s - 15 m / s to 90 degrees, and the prediction accuracy of gas diffusion coverage is increased by 35%; the path is recalculated every 10 seconds - 30 seconds to match the meteorological changes, and the path safety redundancy is increased by 50%.

[0022] 6. Buffer area overlap detection (40%-70% threshold) realizes the fusion of meteorological vectors of multiple leakage points, and the composite sector area of 50 degrees - 120 degrees covers the joint danger range; the hierarchical planning strategy first constructs the main trunk channel and then derives branch paths, reducing the evacuation path conflict in multi-source leakage scenarios by 80% and shortening the overall evacuation time by 25%.

[0023] 7. YOLOv5 identifies the personnel density as 1 person / m² - 3 people / m² and the speed as 0.8 m / s - 1.2 m / s, and accurately marks the congestion nodes; the constraints of slopes of 8 degrees - 15 degrees and detour increments of 120 meters - 200 meters ensure the feasibility of the diversion path; the dynamic road network topology association improves the traffic efficiency of congested sections by 40% and avoids the risk of trampling.

[0024] 8. LoRa beacons (with a radius range of 20 meters to 50 meters) provide centimeter-level positioning of connection nodes. When the deviation of the direction angle is 20 degrees to 40 degrees, a steering prompt is triggered; when deviating from the path for 10 seconds to 20 seconds, vibration alarm + vector map retransmission accelerate the personnel trajectory correction response speed by 3 times; the TTS countdown instruction reduces the steering operation error rate by 60%.

[0025] 9. The gyroscope detects effective steering actions at 15 degrees / second to 30 degrees / second, realizes interactive verification, dynamically extends the countdown by 5 seconds to 10 seconds, and adapts to the operation delay of heavy equipment; vibration mode is differentially adapted, long vibration for heavy equipment and short vibration for light equipment, increasing the alarm perception rate to 98%; noise compensation mechanism, increasing 0.2G - 0.5G for every 10dB - 15dB, ensuring reachability in high-noise environments.

[0026] 10. The standard deviation of the Z-axis acceleration is 0.5G - 8G to accurately determine the violent motion state; the pulse frequency of red or green light is increased by 3Hz - 5Hz or 5Hz - 8Hz to enhance the visual capture rate; the vibration intensity is compensated with motion, increasing 0.1G - 0.3G for every 0.2G - 0.4G, keeping the alarm recognition rate above 90% in the running state; the lumen value increases with the smoke concentration, increasing 30 lumens - 80 lumens for every 100mg / m³ - 200mg / m³, ensuring the visibility of signals in low visibility conditions.

[0027] Other advantages, objectives, and features of the present invention will be partially reflected by the following description and partially understood by those skilled in the art through the research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the composition structure of the intelligent management and control system for chemical industrial park safety risks in one technical solution of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0029] The following further detailed description of the present invention is provided in conjunction with the accompanying drawings, so that those skilled in the art can implement it with reference to the text of the specification.

[0030] It should be understood that terms such as "having", "including", and "comprising" used herein do not exclude the presence or addition of one or more other elements or their combinations.

[0031] According to an embodiment of the present invention, as Figure 1 shown, an intelligent management and control system for chemical industrial park safety risks is provided, including perception layer devices, a data transmission network, a cloud platform data processing center, and application terminals deployed in the chemical industrial park; The perception layer devices include gas concentration sensors, temperature sensors, pressure sensors, and flame detectors arranged in major hazard source tank farms; The data transmission network adopts a dual - mode transmission architecture of 5G communication module and industrial Ethernet. The devices in the perception layer package the collected data and transmit it to the data processing center of the cloud platform. The data processing center of the cloud platform includes a time - series database, a rule engine, and a risk analysis module. The time - series database stores the collected data of the devices in the perception layer in timestamp format and establishes an equipment coding index. The rule engine pre - sets a leakage determination rule: when the methane concentration data associated with the same equipment coding exceeds 25% LEL for 3 to 5 consecutive sampling periods, and the pressure data decline rate is greater than 0.5 MPa / min, a leakage warning is triggered. After receiving the warning signal output by the rule engine, the risk analysis module calls the equipment coordinate data, real - time meteorological data, and emergency resource distribution data in the three - dimensional geographic information system, and calculates the evacuation path within a range of 500 meters to 2000 meters centered on the leakage point through the Dijkstra algorithm. The evacuation path avoids the upwind area of the real - time wind direction and preferentially connects to the fire station coordinate point. The application terminal receives the warning instruction and the evacuation path vector map sent by the data processing center of the cloud platform. The warning instruction includes the leakage equipment coding, warning level, and gas diffusion simulation polygon. The evacuation path vector map includes a dynamically updated set of path node coordinates and the estimated passing time of each node.

[0032] In the devices of the perception layer of this technical solution, the detection range of the gas concentration sensor for methane can be selected from ranges of 0% LEL - 20% LEL, 0% LEL - 50% LEL, or 0% LEL - 100% LEL; the detection range of the temperature sensor can be selected as - 20°C to 150°C, - 40°C to 200°C, or - 60°C to 250°C; the range of the pressure sensor can be selected as 0 MPa - 5 MPa, 0 MPa - 10 MPa, or 0 MPa - 15 MPa; the response wavelength of the flame detector can be selected as 3.3 μm - 4.3 μm, 4.3 μm - 4.5 μm, 4.35 μm - 4.45 μm, or 4.4 μm - 4.6 μm. The time delay of the 5G communication module in the data transmission network is ≤20 ms, preferably less than 5 ms, and optional threshold values are 1 ms, 3 ms, or 5 ms; the transmission bandwidth of the industrial Ethernet is 100 Mbps - 1000 Mbps, and optional values are 100 Mbps, 500 Mbps, or 1000 Mbps. In the leakage determination rule of the rule engine, the methane concentration threshold can be selected as 20% LEL, 25% LEL, or 30% LEL; the consecutive sampling periods can be selected as 3 periods, 4 periods, or 5 periods; the pressure decline rate threshold can be selected as 0.3 MPa / min, 0.5 MPa / min, or 0.7 MPa / min. The calculation radius of the evacuation path can be selected as 500 meters, 1000 meters, or 2000 meters.

[0033] For the devices in the perception layer, commercially available catalytic combustion type methane sensors such as the MC112 model can be selected and installed 1.5 meters downwind of the breathing valve of the storage tank; the PT100 platinum resistance temperature sensor is installed in the middle of the tank wall, 2 to 3 meters above the ground; the piezoresistive pressure sensor such as the MPM4800 model is installed at the pipeline flange interface; the infrared flame detector such as the IR500 model is installed on the angle steel bracket of the tank area cofferdam, 4 to 6 meters above the ground. The 5G communication module of the data transmission network can select an industrial-grade CPE device and be deployed in the explosion-proof control cabinet of the tank area; the industrial Ethernet switch such as the IES618-2GS model is installed in the weak current well of the park. The time series database of the cloud platform data processing center can select InfluxDB and be deployed on the Alibaba Cloud ECS server; the rule engine uses the Drools rule library and runs in the Kubernetes container; the risk analysis module is based on the PostGIS three-dimensional geographic information system and is deployed on an independent GPU computing node. The application terminal can select an explosion-proof tablet computer such as the T8000 model and be equipped at the hanging point of the inspection personnel's protective equipment.

[0034] The working process is as follows: The gas concentration sensor collects methane data every 10 seconds, packs it into the JSON format through the Modbus-RTU protocol, including the timestamp, device code, and concentration value; the temperature and pressure data are sampled at a frequency of 20Hz and uploaded to the cloud platform through a 5G module with a time delay < 3ms or an industrial Ethernet with a bandwidth of 500Mbps. The time series database stores the data indexed by the device code. The rule engine scans the data stream every 30 seconds. When the methane concentration associated with a certain storage tank code is > 25% LEL for 4 consecutive cycles and the pressure drop rate > 0.5MPa / min, a leakage warning is triggered. The risk analysis module calls the real-time wind direction data of the meteorological station, 45° northeast wind, calculates the evacuation path with a radius of 1000 meters centered on the leakage point, avoids the 60° fan-shaped area in the upwind direction, and a distance of 300 meters, and generates the shortest path connecting the coordinates of the nearest fire station X = 102.34, Y = 35.67 through the Dijkstra algorithm. The application terminal receives the warning instruction (including the leakage tank code K-102, high risk level, and the vertex coordinate set of the gas diffusion polygon) and the evacuation path vector map (including 15 path nodes and the passing time of each node). This solution can achieve multi-parameter collaborative warning, improve the reliability of leakage identification; the dual-mode network ensures the integrity of key data transmission; the dynamic path planning reduces the probability of personnel entering the dangerous area.

[0035] According to another embodiment of the present invention, in a chemical industrial park safety risk intelligent management and control system, the risk analysis module further includes a lidar scanning unit, which continuously acquires real-time point cloud data of the path node area after generating the evacuation path; the scanning frequency of the lidar scanning unit is 5Hz - 10Hz, and the scanning angle range covers 30 degrees - 60 degrees on both sides of the evacuation path center line; The risk analysis module performs differential comparison between the real-time point cloud data and the base map elevation data of the three-dimensional geographic information system. When it is identified that the absolute difference between the elevation value of the surface point cloud of the newly added obstacle and the base map elevation data is greater than 50 cm, the evacuation route replanning is triggered; the A-Star algorithm is used in the replanning process to recalculate the obstacle avoidance path, and the updated evacuation route vector map is synchronized to the application terminal within 3 seconds to 8 seconds.

[0036] In this technical solution, the scanning frequency of the lidar scanning unit can be selected as 5Hz, 8Hz or 10Hz; the scanning angle range can be set to cover 30 degrees, 45 degrees or 60 degrees on one side. Among the obstacle recognition thresholds, the absolute elevation difference can be selected as 30 cm, 50 cm or 70 cm; the path replanning response time can be selected as 3 seconds, 5 seconds or 8 seconds. The path node expansion step size of the A-Star algorithm can be set to 1 meter, 2 meters or 3 meters; the obstacle avoidance safety distance threshold can be selected as 0.5 meter, 1 meter or 1.5 meters. The point cloud data sampling interval is 100 milliseconds, 200 milliseconds or 300 milliseconds. The A-Star algorithm is the A* algorithm or A-Star Algorithm.

[0037] The lidar scanning unit can select a 16-line mechanical rotary lidar, which is installed on the top of the street lamp pole in the chemical industrial park, 8 meters to 12 meters above the ground, and the pitch angle is adjusted to the range of -5 degrees to +5 degrees. The point cloud processing module can select an embedded industrial computer equipped with a Jetson TX2 chip, which is deployed in the edge computing box in the park and connected to the lidar through an M12 aviation interface. The base map of the three-dimensional geographic information system is constructed using the Cesium open source engine and runs on the cloud platform GPU server. The application terminal can select an explosion-proof handheld terminal equipped with a 5.7-inch display screen, which is fixed to the buckle of the operator's safety belt.

[0038] The working process is as follows: The lidar scans the 45-degree area on both sides of the center line of the evacuation route at a frequency of 8Hz, with a total coverage of 90 degrees, generating real-time point cloud data, about 30,000 points per frame. The risk analysis module performs differential calculation on the point cloud data and the pre-stored three-dimensional map elevation data, with a grid resolution of 10cm×10cm. When it is detected that the elevation value of the point cloud of a certain grid is 62 cm higher than the base map (corresponding to a collapsed pipe rack), it is determined as a newly added obstacle. After triggering the path replanning, the A-Star algorithm takes the leakage point as the starting point and the fire station coordinates as the ending point, and sets an obstacle avoidance radius of 1.2 meters to recalculate the path (the node expansion step size is 2 meters). The updated path vector map, including the coordinates of 32 path nodes, is sent to the application terminal through the 5G module within 5 seconds. This solution can achieve centimeter-level obstacle recognition and ensure the dynamic availability of the path; the replanning mechanism reduces the risk of path failure caused by environmental mutations.

[0039] According to another embodiment of the present invention, a chemical park safety risk intelligent management and control system is provided, wherein the cloud platform data processing center further includes an instruction distribution module connected to the risk analysis module and the application terminal, and the instruction distribution module includes: The terminal positioning unit obtains physical location data through the GPS positioning module built into the application terminal, with a positioning accuracy error of less than 3 meters; The distance classification unit calculates the straight-line distance between each application terminal and the leakage point and divides it into three priority levels. Among them, the distance to the leakage point is less than or equal to 500 meters, the distance to the leakage point is greater than 500 meters and less than or equal to 1000 meters, the distance to the leakage point is greater than 1000 meters and less than or equal to 2000 meters, the distance to the leakage point is greater than 2000 meters, and the warning is automatically lifted. When the evacuation path re-planning is triggered, the transmission control unit directly transmits the evacuation path vector diagram to the first-priority application terminal, and sends it through the 5G communication module using the UDP protocol. The data packets transmitted to the second and third-priority application terminals are stored in the Redis cache queue and sent through the industrial Ethernet after a delay of 2 to 5 seconds. The resource allocation ratio of the transmission control unit is dynamically adjusted according to the personnel evacuation priority strategy in the chemical park emergency plan, among which the first-priority application terminal allocates 60% to 80% of the total bandwidth, and the remaining broadband is allocated according to the ratio of the number of second-priority application terminals to the number of third-priority application terminals = 3:1.

[0040] In this technical solution, the GPS positioning accuracy error threshold of the terminal positioning unit can be set to 1 meter, 2 meters or 3 meters. In the distance classification unit, the first priority distance threshold can be selected as 300 meters, 500 meters or 800 meters; the second priority can be selected from 800 meters-1200 meters, 500 meters-1000 meters or 1000 meters-1500 meters; the third priority warning release distance threshold can be selected from 1500 meters, 2000 meters or 2500 meters. The direct transmission delay of the transmission control unit can be selected from 100ms, 200ms or 300ms; the Redis cache queue delay time can be selected from 2 seconds, 3 seconds or 5 seconds; the first priority bandwidth share can be selected from 60%, 70% or 80%; the second and third priority allocation ratios can be selected from 2:1, 3:1 or 4:1. The "resources" in this technical solution specifically refer to the network bandwidth resources managed by the instruction distribution module. The total bandwidth is dynamically allocated by the cloud platform based on the current number of concurrent connections. The benchmark value is 1Gbps-2Gbps, and the total resources are network broadband.

[0041] The terminal positioning unit can utilize a multi-frequency GNSS positioning module supporting GPS / Beidou, built into the explosion-proof application terminal motherboard, with the antenna located beneath the terminal's plastic housing. The distance classification unit can utilize an edge computing gateway equipped with an i.MX8 processor, deployed on the second floor of a 19-inch cabinet in the campus communications room. The transmission control unit's Redis cache server can utilize an in-memory database cluster, installed in a cloud platform virtualization environment. The 5G communication module can utilize an industrial-grade CPE device, fixed 1.5 meters above the ground to an explosion-proof column in the tank area. A 24-port Gigabit industrial Ethernet switch can be installed in the control center's wiring closet. The application terminal housing can be constructed from a flame-retardant PC / ABS alloy, with the internal circuit boards coated with conformal coating.

[0042] The operating process is as follows: When the risk analysis module triggers path replanning, the terminal positioning unit uses GPS to obtain the location of a terminal (longitude 118.76°, latitude 32.04°). The distance classification unit calculates the straight-line distance from the leak point (longitude 118.75°, latitude 32.05°) as 480 meters, assigning it the first priority. The transmission control unit directly transmits the path vector diagram (containing 28 nodes) to the terminal via UDP, with 5G transmission latency controlled to 120ms. Simultaneously, the data packet from the second-priority terminal (780 meters away) is stored in a Redis queue and distributed via Industrial Ethernet three seconds later. Bandwidth allocation is dynamically adjusted to: 70% of the first-priority terminal's available bandwidth (1.4 Gbps), and the remaining 30% is allocated proportionally to the number of terminals: 22.5% for the 15 second-priority terminals and 7.5% for the five third-priority terminals. This solution ensures that terminals in high-risk areas receive priority path updates, alleviating network congestion. The tiered delay mechanism optimizes resource utilization.

[0043] According to another embodiment of the present invention, a method for intelligently managing and controlling safety risks in a chemical park is provided, comprising the following steps: Step 1: Real-time methane concentration data is collected through gas concentration sensors deployed in major hazardous source tank areas; Step 2: collecting tank area temperature data through a temperature sensor; Step 3: Collecting tank area pressure data through pressure sensors; Step 4: Collect flame radiation data through a flame detector; Step 5: Send the methane concentration data, temperature data, pressure data, and flame radiation data collected in steps 1 to 4 to the cloud platform data processing center via the 5G communication module and industrial Ethernet dual-mode transmission architecture; Step 6: Store the received data in the time series database of the cloud platform data processing center and create a device code index in a timestamp format; Step 7, execute leakage determination through the rule engine: When the methane concentration data associated with the same device code exceeds 25% LEL for 3 to 5 consecutive sampling periods, and the pressure data decline rate is greater than 0.5 MPa / min, a leakage warning signal is generated; Step 8, in response to the leakage warning signal, call the device coordinate data, real-time meteorological data, and emergency resource distribution data stored in the 3D geographic information system; Step 9, set the evacuation route avoidance area based on the real-time wind direction data. The evacuation route area covers the 30° - 60° fan-shaped area upwind of the leakage point, within a range of 200 meters to 800 meters from the leakage point; Step 10, calculate the evacuation route within a range of 500 meters to 2000 meters centered on the leakage point through the Dijkstra algorithm. The evacuation route bypasses the evacuation route avoidance area set in Step 9 and is preferentially connected to the nearest fire station coordinate point; Step 11, encapsulate the leakage device code, warning level, and gas diffusion simulation polygon into a warning instruction; Step 12, encapsulate the evacuation route node coordinate set, the estimated passing time of each node, and the real-time wind direction vector into an evacuation route vector map; Step 13, send the warning instruction and the evacuation route vector map to the application terminal through the 5G communication module.

[0044] In Step 1 of this technical solution, the methane concentration detection range can be selected from 0% LEL - 50% LEL, 0% LEL - 100% LEL, or 0% LEL - 150% LEL ranges. In Step 2, the temperature detection range can be selected from -30°C to 150°C, -40°C to 200°C, or -50°C to 250°C. In Step 3, the pressure detection range can be selected from 0 MPa - 8 MPa, 0 MPa - 10 MPa, or 0 MPa - 12 MPa. In Step 4, the response wavelength of the flame detector can be selected from 3.3 μm - 4.3 μm, 4.3 μm - 4.5 μm, 4.35 μm - 4.45 μm, or 4.4 μm - 4.6 μm. In Step 5, the 5G transmission delay can be selected from 3 ms, 5 ms, or 8 ms; the industrial Ethernet bandwidth can be selected from 100 Mbps, 500 Mbps, or 1000 Mbps. In Step 7, the continuous period for leakage determination can be selected from 3 periods, 4 periods, or 5 periods; the methane threshold can be selected from 20% LEL, 25% LEL, or 30% LEL; the pressure decline rate threshold can be selected from 0.4 MPa / min, 0.5 MPa / min, or 0.6 MPa / min. In Step 9, the avoidance area angle can be selected from 30°, 45°, or 60°; the distance can be selected from 200 meters, 500 meters, or 800 meters. In Step 10, the evacuation radius can be selected from 500 meters, 1000 meters, or 2000 meters. In Step 13, the sending delay can be selected from 500 ms, 800 ms, or 1000 ms.

[0045] The sensors in Steps 1 to 4 can select the same type of equipment in the foregoing technical solutions: the catalytic combustion type methane sensor is installed at the flange interface 1.5 meters downwind of the storage tank breathing valve; the PT100 temperature sensor is assembled on the side wall of the tank body at a bracket 2 meters above the ground; the piezoresistive pressure sensor is connected to the pipeline blowdown valve interface; the infrared flame detector is fixed at the angle steel column of the tank area cofferdam 4.5 meters above the ground. The 5G communication module in Step 5 can select an industrial-grade CPE and is deployed in the explosion-proof control box of the tank area; the industrial Ethernet switch is installed in the weak current cabinet of the park. The time series database in Step 6 can select InfluxDB and run on an Alibaba Cloud 4-core 8G ECS server. The three-dimensional geographic information system in Step 8 can select the SuperMap platform and be deployed on an independent GPU server. The application terminal can select an explosion-proof tablet computer and be equipped on the operator's safety belt. Step 8 also includes calculating the gas diffusion range based on the Gaussian plume diffusion model. The input parameters include: the leakage point coordinates, real-time wind speed data, wind direction angle data, and leakage duration. The output is the diffusion concentration isopleth; the closed area of the isopleth with a concentration ≥ 25% LEL is converted into polygon vertex coordinates and stored as a gas diffusion simulation polygon.

[0046] The working process is as follows: The temperature sensor collects the tank wall temperature data every 15 seconds (the current value is 65 °C), and the pressure sensor monitors the pipeline pressure at a frequency of 20 Hz (the current value is 3.2 MPa). The data is packaged into JSON format through the Modbus-RTU protocol. It is uploaded to the cloud platform through a 5G module with a time delay of 4 ms, and the time series database stores and indexes it according to the device code. The rule engine scans the data every 30 seconds. When the methane concentration associated with the storage tank B-203 code is continuously > 25% LEL for 4 times and the pressure drop rate > 0.5 MPa / min, a leakage warning is triggered. The risk analysis module calls the real-time northeast wind data, demarcates an avoidance area of 45 degrees upwind and 500 meters away from the leakage point, and generates an evacuation path with a radius of 1000 meters through the Dijkstra algorithm (bypassing the avoidance area and connecting to the coordinates of the fire station X102 / Y58). The warning instruction includes the leakage code, high risk level, and gas diffusion polygon coordinates; the evacuation path vector map includes 18 nodes and the passing time of each node, and is sent to the terminal within 720 ms through the 5G module. This solution realizes the automation of leakage determination and improves the emergency response speed; the dynamic path avoidance reduces the risk of personnel exposure.

[0047] According to another embodiment of the present invention, in the intelligent control method for chemical industrial park safety risks, step eight further includes: extracting wind speed data and wind direction angle data from real-time meteorological data, where the wind speed detection range is 1 m / s - 30 m / s, and the wind direction angle detection range is 0 degrees to 360 degrees; performing cubic spline interpolation calculation on the wind speed data and wind direction angle data with an interpolation period of 15 seconds - 40 seconds to generate a wind field model with a continuous time series; associating the output of the wind field model with the leakage point equipment code and storing it as dynamic meteorological vector data; The path avoidance area set in step nine is updated based on the dynamic meteorological vector data. When the wind speed data is 8 m / s - 15 m / s, the fan-shaped area angle expands to the range of 45 degrees - 90 degrees; The calculation frequency of the evacuation path in step ten is increased to be executed once every 10 seconds - 30 seconds.

[0048] In step eight of this technical solution, the wind speed detection range can be selected from ranges of 1 m / s - 10 m / s, 1 m / s - 20 m / s, or 1 m / s - 30 m / s; the wind direction angle detection uses full coverage of 0° - 360°. The interpolation period can be selected as 15 seconds, 25 seconds, or 40 seconds; the number of cubic spline interpolation nodes can be set to 4 points, 6 points, or 8 points. In the angle adjustment threshold of the avoidance area in step nine, the wind speed trigger value can be selected as 6 m / s, 8 m / s, or 10 m / s; the angle expansion range can be selected as 45° - 60°, 60° - 75°, or 75° - 90°. The path calculation frequency in step ten can be selected to be executed every 10 seconds, 20 seconds, or 30 seconds.

[0049] The wind speed and direction sensor can select an ultrasonic weather station, which is installed on the roof lightning rod base of the tallest building in the chemical industrial park, with a height from the ground not less than 20 meters. The interpolation calculation module can select a cloud platform GPU server, equipped with an NVIDIA T4 computing card, and deployed on the third layer of the data center rack. The dynamic meteorological vector storage unit can select a Redis time series database, running in a 16GB memory container. The three-dimensional geographic information system is implemented using SuperMap platform software and is connected to the meteorological data interface through the TCP / IP protocol. The application terminal display screen can select an explosion-proof liquid crystal panel made of Gorilla Glass and is embedded in the terminal housing.

[0050] The working process is as follows: The ultrasonic weather station collects the wind speed (current value: 12 m / s) and wind direction (current value: 135°) every 2 seconds and uploads them to the cloud platform through the OPC protocol. Taking 25 seconds as the interpolation period, cubic spline interpolation is performed on the wind speed and wind direction data of the nearest 4 cycles to generate a continuous wind field model (including wind speed vectors at 5° intervals in the direction of 0° - 360°). When the wind speed value remains > 8 m / s for 30 seconds, in Step Nine, the avoidance area angle is extended from 60° to 80°. The risk analysis module calls the updated dynamic meteorological vector every 20 seconds to recalculate the evacuation route (for example, if the original route node N15 is located in the diffusion area, it is adjusted to the downwind safe node N16). The path vector map contains wind direction arrow markings and is sent down through the 5G module. This solution enables the dynamic adjustment of the avoidance area according to meteorological changes, enhancing the path safety; the high-frequency recalculation adapts to sudden meteorological changes.

[0051] According to another embodiment of the present invention, in the intelligent control method for safety risks in chemical industrial parks, Step Eight further includes: retrieving the device codes associated with all current leakage warning signals, extracting the coordinate data of each leakage point and the dynamic meteorological vector data; establishing a buffer area with a radius of 30 meters - 60 meters centered on the leakage point coordinates, and using a spatial overlay analysis algorithm to detect the overlapping part of the buffer area; when the overlapping area ratio exceeds 40% - 70%, merging the dynamic meteorological vector data and generating a combined path avoidance area; In Step Nine, the set range of the path avoidance area is extended to: maintaining a 30-degree - 60-degree sector for a single leakage point; adjusting to a 50-degree - 120-degree composite sector for the combined path avoidance area, and the distance is taken as the maximum of 200 meters - 800 meters among the leakage points; In Step Ten, the evacuation route calculation adopts a hierarchical planning strategy: first, plan the main evacuation channels to connect the safety exit coordinate points within a radius of 500 meters - 2000 meters; then generate branch paths based on the main evacuation channels to each device code area.

[0052] In Step Eight of this technical solution, the buffer area radius can be selected as 30 meters, 45 meters, or 60 meters; the overlapping area ratio threshold can be selected as 40%, 55%, or 70%. For the composite sector angle in Step Nine, it can be selected as 50° - 80°, 80° - 100°, or 100° - 120°; the distance range is taken as the maximum of 200 meters - 400 meters, 400 meters - 600 meters, or 600 meters - 800 meters among the leakage points. In the hierarchical planning of Step Ten, the connection radius of the main channel safety exits can be selected as 500 meters - 1000 meters, 1000 meters - 1500 meters, or 1500 meters - 2000 meters; the branch path generation granularity can be selected as 50-meter, 100-meter, or 150-meter grids.

[0053] The spatial overlay analysis module can select the ArcGIS Engine component and be deployed on the spatial computing server of the cloud platform. The buffer area generation unit can select the PostGIS geodatabase extension and run on a 128GB memory database node. The dynamic meteorological vector memory is implemented using a Redis time series database. The backbone channel planning server can select a highly available cluster (dual Xeon Gold processors) and be installed on the 5th layer of the data center rack. The video monitoring device can select an explosion-proof dome camera with a stainless steel housing and be installed on a bracket 6 meters above the ground in the pipe gallery of the tank area. The merging rule for merging dynamic meteorological vector data can be: take the maximum wind speed among the wind speeds at each leakage point as the combined wind speed, and take the wind direction at the leakage point with the maximum wind speed as the dominant wind direction; if the leakage points with the maximum wind speed are not unique, then take the wind direction at the geometric center point of the coordinates.

[0054] The working process is as follows: When the storage tank area T-201 (coordinates X = 102.35, Y = 35.68) and the distillation column R-305 (coordinates X = 102.41, Y = 35.72) simultaneously trigger a leakage warning, in step eight, buffer circles with a radius of 50 meters are established centered at each point. The spatial overlay analysis detects that the overlapping area ratio of the two circles reaches 58% (>55% threshold), and the dynamic meteorological vectors are merged to generate a combined avoidance area. In step nine, the angle of the composite sector area is set to 100°, and the distance is taken as the maximum value of 600 meters (covering the 180-meter gap area between the two leakage points). In step ten, first, the backbone evacuation channel is planned to connect to the west No. 2 safety exit (800 meters away from the T-201 point), and then branch paths are regenerated to the operating platform of the T-201 tank area (including 3 path nodes) and the control room of the R-305 (including 2 path nodes). This solution can handle the coupling risk of multiple leakage points, avoid path crossing conflicts; and optimize the calculation efficiency through hierarchical planning.

[0055] According to another embodiment of the present invention, in the intelligent management and control method for safety risks in chemical industrial parks, the branch path generation stage in step ten includes the following steps: S1. Obtain the real-time video monitoring data along the backbone evacuation channel, and identify the personnel density and moving speed in the backbone evacuation channel through the YOLOv5 target detection algorithm; S2. When the personnel density is 1 person / square meter - 3 people / square meter and the moving speed is 0.8 m / s - 1.2 m / s, mark the section of the backbone evacuation channel as a congestion node; S3. Retrieve the alternative branch paths within a range of 50 meters - 100 meters from the congestion node, and calculate the detour distance increment and slope data of the alternative branch paths; S4. Select the alternative branch path with a detour distance increment of 120 meters - 200 meters and a slope of 8 degrees - 15 degrees as the diversion path; S5. Topologically associate the coordinates of the diversion path nodes with the nodes of the backbone evacuation channel to generate a dynamic evacuation road network; The evacuation path vector map encapsulation in Step 12 contains topological relation data of the dynamic evacuation road network.

[0056] In Step S1 of this technical solution, the personnel density detection threshold can be selected as 1 person / square meter, 2 persons / square meter, or 3 persons / square meter; the moving speed threshold can be selected as 0.8 m / s, 1.0 m / s, or 1.2 m / s. The combination of congestion node marking conditions in Step S2 is: density 1 person / ㎡ - 2 persons / ㎡ and speed 0.8 m / s - 1.0 m / s, or density 2 persons / ㎡ - 3 persons / ㎡ and speed 1.0 m / s - 1.2 m / s. In Step S3, the standby path retrieval radius can be selected as 50 meters, 75 meters, or 100 meters; the detour distance increment threshold is set as 120 meters, 160 meters, or 200 meters; the slope threshold can be selected as 8°, 12°, or 15°. The constraint condition for the diversion path selection in Step S4 is: simultaneously satisfying the detour increment ≤ 180 meters and the slope ≤ 12°.

[0057] The video surveillance device can select a 2-million-pixel stainless steel 316L shell explosion-proof network camera or a 4K resolution camera, which is installed on the pipe gallery support above the main evacuation passage, with a height from the ground not exceeding 6 meters and a depression angle of 15°. The target detection server can select an edge computing box equipped with a Jetson Xavier NX chip and is deployed in the explosion-proof control box on the side wall of the passage. The slope detection sensor can select a MEMS inclinometer (range ±30°) and is fixed on the back of the branch path sign. The dynamic evacuation road network generation module is implemented using the ArcGISEngine spatial analysis component and runs on the spatial analysis server of the cloud platform. The application terminal display screen can select a high-brightness IPS liquid crystal panel with surface anti-reflection treatment and is embedded in the front of the terminal shell.

[0058] The working process is as follows: When the video surveillance in Area C of the main passage shows that the personnel density reaches 2.3 persons / square meter and the average moving speed identified by the YOLOv5 algorithm is 0.9 m / s, Step S2 marks this section as a congestion node. Step S3 retrieves 3 standby branch paths within a radius of 80 meters from the node: Path A has a detour increment of 150 meters / slope of 10°, Path B has a detour increment of 220 meters / slope of 8°, and Path C has a detour increment of 180 meters / slope of 16°. Step S4 selects Path A (satisfying the increment ≤ 180 meters and the slope ≤ 12°). Step S5 topologically associates the nodes of Path A (N45 - N48) with the node N32 of the main passage to generate dynamic evacuation road network vector data. The application terminal receives the updated path map and displays the diversion arrow mark. This solution can relieve the congestion of the main passage and ensure the smoothness of evacuation; the slope constraint ensures the feasibility of the path.

[0059] According to another embodiment of the present invention, in the intelligent control method for chemical industrial park safety risks, step twelve further includes: parsing the topological relationship data of the dynamic evacuation road network, and extracting the coordinates of the connection nodes between the diversion paths and the main evacuation channels; deploying LoRa wireless beacons within a range of 20 meters to 50 meters from the connection nodes, with the beacon broadcast frequency being 1 time per second to 3 times per second, and the broadcast content including node codes and path direction angles; after the application terminal receives the LoRa beacon data, comparing it with the dynamic evacuation road network vector map stored locally: if the distance between the current position of the application terminal and the connection node is 15 meters to 30 meters and the direction angle deviation is 20 degrees to 40 degrees, trigger a voice turning prompt; if the moving trajectory of the application terminal deviates from the dynamic evacuation road network planned path for 10 seconds to 20 seconds continuously, start vibration warning and retransmit the dynamic evacuation road network vector map that is newly generated and updated in real time by the cloud platform data processing center to this application terminal; the voice prompt content is generated by using a TTS engine to convert path instructions, including the turning angle value and a countdown within a range of 15 seconds to 30 seconds.

[0060] In this technical solution, the deployment radius of the connection nodes can be selected as 20 meters, 35 meters or 50 meters; the beacon broadcast frequency can be selected as 1 time per second, 2 times per second or 3 times per second. The direction angle deviation threshold is set as 20°, 30° or 40°; the deviation path duration threshold can be selected as 10 seconds, 15 seconds or 20 seconds. The TTS countdown range can be selected as 15 seconds, 22 seconds or 30 seconds. The accuracy of the voice prompt turning angle value is ±5°.

[0061] The LoRa wireless beacon can select an industrial-grade micro-power consumption module (IP68 enclosure), and is installed in the middle of the street lamp pole at the connection of the branch path and the main channel, 2.5 meters above the ground. The gyroscope module can select a MEMS six-axis sensor (range ±2000° / second) and is built into the main board of the application terminal. The accelerometer uses a three-axis MEMS sensor. The TTS engine can select an offline voice synthesis chip that supports Chinese instructions and is integrated beside the terminal processor. The vibration motor can select a flat eccentric rotor motor (diameter 10mm) and is fixed inside the terminal shell.

[0062] The working process is as follows: when the application terminal approaches the connection node (coordinates X = 102.56, Y = 35.73), the LoRa beacon broadcasts the node code N58 and the direction angle 285° at a frequency of 2Hz. The terminal compares the local path map, detects that the current position is 25 meters away from the node and the direction angle deviation is 35° (>30° threshold), and triggers a voice prompt: "Turn left 40 degrees, countdown 20 seconds". If the moving trajectory of the terminal deviates from the planned path for 15 seconds continuously (such as moving in a zigzag), start vibration warning and retransmit the updated vector map (including the newly added diversion node N59). The vibration intensity increases to 2.2G with the ambient noise (current 85dB). This solution realizes centimeter-level path correction and reduces the risk of personnel getting lost; multi-modal warnings adapt to complex environments.

[0063] According to another embodiment of the present invention, in the intelligent control method for safety risks in chemical industrial parks, step twelve further includes: when the voice steering prompt is triggered, the gyroscope and accelerometer of the application terminal are synchronously activated, and the sampling frequency is 50Hz - 100Hz; the angular velocity value of the steering action of the application terminal holder is calculated in real time, and the effective steering threshold is set to 15 degrees / second - 30 degrees / second; If no effective steering action is detected within 3 seconds - 8 seconds after the countdown starts, the remaining countdown is extended by 5 seconds - 10 seconds; if the cumulative extension times reach 2 times - 4 times, the steering voice prompt is turned off and the vibration alarm is started; The vibration alarm mode is adjusted according to the type of protective equipment bound to the application terminal: when the protective equipment is a heavy-duty chemical protection suit, 3 times - 5 times of long vibrations lasting for 1 second - 3 seconds are adopted; when the protective equipment is a light protective suit, 8 times - 12 times of short vibrations lasting for 0.3 seconds - 0.6 seconds are adopted; the vibration intensity linearly increases with the environmental noise decibel value, and the reference intensity is 1.5G - 2.5G, and 0.2G - 0.5G is increased for every 10 decibels - 15 decibels increase in noise.

[0064] In this technical solution, the sampling frequency of the gyroscope can be selected as 50Hz, 75Hz or 100Hz; the effective steering threshold is set to 15° / second, 22° / second or 30° / second. The first extension amount of the countdown can be selected as 5 seconds, 7 seconds or 10 seconds; the cumulative extension times threshold is set to 2 times, 3 times or 4 times. The long vibration mode can be selected as 3 times lasting for 1.5 seconds, 4 times lasting for 2 seconds or 5 times lasting for 3 seconds; the short vibration mode can be selected as 8 times lasting for 0.4 seconds, 10 times lasting for 0.5 seconds or 12 times lasting for 0.6 seconds. The noise compensation reference intensity can be selected as 1.5G, 2.0G or 2.5G; 0.2G, 0.3G or 0.5G is increased for every 10dB increase in noise.

[0065] The gyroscope module can select a MEMS angular velocity sensor (±4000dps range), which is co-welded with the accelerometer on the terminal PCB. The vibration motor can select a linear resonator actuator (LRA) or an eccentric rotor motor (ERM), which is glued to the side wall of the terminal battery compartment. The protective equipment binding unit can select an anti-metal NFC tag, which is sewn inside the right shoulder strap of the chemical protection suit. The noise sensor can select a digital microphone module (frequency response 100Hz - 10kHz), which is embedded in the opening at the top of the terminal.

[0066] The working process is as follows: After the operator wearing a heavy chemical protective suit receives a steering prompt, the gyroscope samples and detects the angular velocity at 80 Hz (the current value is 18° / second < the threshold of 22° / second). If the steering threshold is not reached after the countdown starts for 5 seconds, the countdown is extended by 7 seconds. If the effective steering is still not detected after 3 consecutive extensions, the voice prompt is turned off and the long vibration mode is activated (4 times for 2 seconds each, with an intensity of 2.8 G). The vibration intensity increases to 3.3 G with the compressor noise (95 dB) (the reference is 2.5 G + 0.8 G compensation). This solution adapts to the operation delay of heavy equipment to ensure that key instructions are reached; the intensity compensation mechanism maintains the warning perception.

[0067] According to another embodiment of the present invention, in the intelligent control method for the safety risks in the chemical industrial park, step twelve further includes: real-time collecting the three-axis acceleration data through the accelerometer of the application terminal, and the sampling frequency is maintained at 50 Hz - 100 Hz; when the standard deviation of the Z-axis acceleration exceeds 0.5 G - 8 G within 3 seconds - 5 seconds continuously, it is determined that the application terminal is in a violent motion state; In the violent motion state, the vibration warning and the LED indication module are activated synchronously: when the protective equipment is a heavy chemical protective suit, the red light pulse frequency is increased to 3 Hz - 5 Hz, and the vibration intensity is increased by 0.5 G - 1.0 G; when the protective equipment is a light protective suit, the green light pulse frequency is increased to 5 Hz - 8 Hz, and the vibration mode is switched to intermittent strong vibration, with a working cycle of vibrating once every 0.2 seconds to 0.5 seconds, and the interval between each vibration is 0.1 seconds to 0.3 seconds; The vibration intensity compensation value is dynamically adjusted according to the motion intensity, including: the reference intensity is the set value during the adaptation stage of the protective equipment; for every 0.2 G - 0.4 G increase in the standard deviation of the Z-axis acceleration, the vibration intensity is increased by 0.1 G - 0.3 G; The brightness of the light pulse increases with the smoke concentration. For every 100 mg / m³ - 200 mg / m³ increase in the smoke concentration, the LED output lumen value is increased by 30 lumens - 80 lumens.

[0068] The sampling frequency of the accelerometer in this technical solution is maintained at 50 Hz, 75 Hz or 100 Hz. The violent motion determination threshold: the standard deviation of the Z-axis acceleration can be 0.5 G, 4.0 G or 8.0 G; the continuous duration can be 3 seconds, 4 seconds or 5 seconds. The red light pulse frequency can be 3 Hz, 4 Hz or 5 Hz; the green light pulse frequency can be 5 Hz, 6 Hz or 8 Hz. The vibration intensity compensation value: for every 0.2 G standard deviation increase, it is increased by 0.1 G, for every increase of 0.3 G it is increased by 0.2 G, or for every increase of 0.4 G it is increased by 0.3 G. Smoke compensation: for every 100 mg / m³ increase, it is increased by 30 lumens, for every increase of 150 mg / m³ it is increased by 55 lumens, or for every increase of 200 mg / m³ it is increased by 80 lumens.

[0069] An industrial-grade MEMS module (±16g range) can be selected for the three-axis accelerometer and mounted at the center of the terminal main board. A high-brightness RGB LED (200 lumens) can be selected for the LED indication module and embedded under the transparent protective cover at the top of the terminal. A laser scattering type probe (range 0mg / m³ - 500mg / m³) can be selected for the smoke sensor and installed behind the intake grille on the side of the terminal. The vibration motor uses an eccentric rotor motor and is connected in parallel with the LED synchronous control circuit.

[0070] The working process is as follows: When a person runs, the accelerometer samples at 90Hz, and the standard deviation of the Z-axis reaches 6.2G for 4 consecutive seconds (> 4.0G threshold), determining the violent motion state. When wearing light protective clothing: Activate the green light pulse (6Hz frequency) and switch to the intermittent strong vibration mode (vibrate once every 0.3 seconds / interval 0.2 seconds). The vibration intensity is increased from the reference 1.8G to 2.4G (compensation value = 6.2G / 0.3 × 0.15). At the same time, the ambient smoke concentration is detected as 220mg / m³, and the LED lumen value is increased from 150 lumens to 260 lumens (reference value + 110 lumen compensation). This solution enhances the alarm recognition rate in the motion state; the light-vibration linkage compensation mechanism improves the reliability of perception in harsh environments.

[0071] The number of devices and the processing scale described here are used to simplify the description of the present invention. Applications, modifications, and variations of the present invention will be apparent to those skilled in the art.

[0072] Although the embodiments of the present invention have been disclosed as above, it is not limited to only the applications listed in the specification and the embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to specific details and the illustrated and described examples here.

Claims

1. The intelligent control system for safety risks in chemical industrial parks is characterized in that It includes sensing layer devices, a data transmission network, a cloud platform data processing center, and application terminals deployed in a chemical industrial park; The sensing layer devices include gas concentration sensors, temperature sensors, pressure sensors, and flame detectors installed in major hazard source tank farms; The data transmission network adopts a dual-mode transmission architecture of 5G communication modules and industrial Ethernet. The sensing layer devices package the collected data and transmit it to the cloud platform data processing center; The cloud platform data processing center includes a time series database, a rule engine, and a risk analysis module. The time series database stores the collected data of the sensing layer devices in timestamp format and establishes an equipment coding index. The rule engine pre-sets a leakage determination rule: when the methane concentration data associated with the same equipment coding exceeds 25% LEL for 3 to 5 consecutive sampling periods, and the pressure data decline rate is greater than 0.5 MPa / min, a leakage warning is triggered. After receiving the warning signal output by the rule engine, the risk analysis module calls the equipment coordinate data, real-time meteorological data, and emergency resource distribution data in the three-dimensional geographic information system, and calculates the evacuation path within a range of 500 meters to 2000 meters centered on the leakage point through the Dijkstra algorithm. The evacuation path avoids the upwind area of the real-time wind direction and preferentially connects to the fire station coordinate point; The application terminal receives the warning instruction and the evacuation path vector map sent by the cloud platform data processing center. The warning instruction includes the leakage equipment coding, warning level, and gas diffusion simulation polygon. The evacuation path vector map includes a dynamically updated path node coordinate set and the estimated passing time of each node; 2. The intelligent control system for chemical industrial park safety risks as claimed in claim 1, wherein The risk analysis module also includes a lidar scanning unit, which continuously obtains the real-time point cloud data of the path node area after generating the evacuation path. The scanning frequency of the lidar scanning unit is 5 Hz to 10 Hz, and the scanning angle range covers 30 degrees to 60 degrees on both sides of the evacuation path center line; The risk analysis module performs differential comparison between the real-time point cloud data and the base map elevation data of the three-dimensional geographic information system. When it is identified that the absolute difference between the elevation value of the surface point cloud of the newly added obstacle and the base map elevation data is greater than 50 cm, an evacuation path replanning is triggered. The A-Star algorithm is used in the replanning process to recalculate the obstacle avoidance path, and the updated evacuation path vector map is synchronized to the application terminal within 3 seconds to 8 seconds; 3. The intelligent management and control system for chemical industrial park safety risks according to claim 2, wherein, The cloud platform data processing center also includes an instruction distribution module connected to both the risk analysis module and the application terminal. The instruction distribution module includes: A terminal positioning unit that obtains physical location data through the GPS positioning module built in the application terminal, and the positioning accuracy error is less than 3 meters; A distance grading unit that calculates the straight-line distance between each application terminal and the leakage point and divides it into 3 priority levels. Among them, the distance less than or equal to 500 meters from the leakage point is the first priority, the distance greater than 500 meters and less than or equal to 1000 meters from the leakage point is the second priority, the distance greater than 1000 meters and less than or equal to 2000 meters from the leakage point is the third priority, and the warning is automatically cancelled when the distance is greater than 2000 meters; The transmission control unit, when the evacuation route replanning is triggered, directly transmits the evacuation route vector map to the first-priority application terminal, which is sent through the 5G communication module using the UDP protocol. The data packets transmitted to the second- and third-priority application terminals are stored in the Redis cache queue and sent through the industrial Ethernet after a delay of 2 to 5 seconds. The resource allocation ratio of the transmission control unit is dynamically adjusted according to the personnel evacuation priority strategy in the chemical industrial park emergency plan. Among them, the resource allocation ratio of the first-priority application terminal accounts for 60% - 80% of the total bandwidth, and the remaining bandwidth is allocated according to the ratio of the number of second-priority application terminals: the number of third-priority application terminals = 3:

1.

4. Intelligent control method for safety risks in chemical industrial parks, characterized in that, It includes the following steps: Step 1, collect methane concentration data in real time through gas concentration sensors deployed in the major hazard source tank area; Step 2, collect tank area temperature data through temperature sensors; Step 3, collect tank area pressure data through pressure sensors; Step 4, collect flame radiation data through flame detectors; Step 5, send the methane concentration data, temperature data, pressure data, and flame radiation data collected in Steps 1 to 4 to the cloud platform data processing center through the 5G communication module and the industrial Ethernet dual-mode transmission architecture; Step 6, store the received data in the time series database of the cloud platform data processing center and establish a device coding index in the timestamp format; Step 7, execute leakage determination through the rule engine: when the methane concentration data associated with the same device coding exceeds 25% LEL for 3 to 5 consecutive sampling periods and the pressure data decline rate is greater than 0.5 MPa / min, generate a leakage warning signal; Step 8, in response to the leakage warning signal, call the device coordinate data, real-time meteorological data, and emergency resource distribution data stored in the three-dimensional geographic information system; ​ ​ ​ ​ ​ 5. The intelligent control method for safety risks in chemical industrial parks according to claim 4, characterized in that, ​ The path avoidance area set in Step Nine is updated based on dynamic meteorological vector data. When the wind speed data is 8 m / s - 15 m / s, the fan-shaped area angle expands to the range of 45 degrees - 90 degrees; In Step Ten, the calculation frequency of the evacuation path is increased to be executed once every 10 seconds - 30 seconds.

6. The intelligent control method for safety risks in chemical industrial parks according to claim 5, wherein Step Eight also includes: retrieving the device codes associated with all current leakage warning signals, extracting the coordinate data of each leakage point and the dynamic meteorological vector data; establishing a buffer area with a radius of 30 m - 60 m centered on the leakage point coordinates, and using the spatial overlay analysis algorithm to detect the overlapping parts of the buffer area; when the overlapping area ratio exceeds 40% - 70%, merging the dynamic meteorological vector data and generating a combined path avoidance area; The set range of the path avoidance area in Step Nine is extended to: maintaining a 30-degree - 60-degree fan-shaped area for a single leakage point; adjusting to a 50-degree - 120-degree composite fan-shaped area for the combined path avoidance area, and the distance is taken as the maximum of 200 m - 800 m among the leakage points; In Step Ten, the hierarchical planning strategy is adopted for calculating the evacuation path: first, plan the main evacuation channels to connect the coordinate points of the safety exits within a radius of 500 m - 2000 m; then generate branch paths based on the main evacuation channels to each device code area.

7. The intelligent control method for safety risks in chemical industrial parks according to claim 6, wherein, The branch path generation stage in Step Ten includes the following steps: S1. Obtain the real-time video surveillance data along the main evacuation channels, and identify the personnel density and moving speed in the main evacuation channels through the YOLOv5 object detection algorithm; S2. When the personnel density is 1 person / m² - 3 persons / m² and the moving speed is 0.8 m / s - 1.2 m / s, mark this section of the main evacuation channel as a congestion node; S3. Retrieve the alternative branch paths within a radius of 50 m - 100 m of the congestion node, and calculate the detour distance increment and slope data of the alternative branch paths; S4. Select the alternative branch path with a detour distance increment of 120 m - 200 m and a slope of 8 degrees - 15 degrees as the diversion path; S5. Topologically associate the node coordinates of the diversion path with the main evacuation channel nodes to generate a dynamic evacuation road network; The encapsulation of the evacuation path vector map in Step Twelve contains the topological relationship data of the dynamic evacuation road network.

8. The intelligent control method for safety risks in chemical industrial parks according to claim 7, wherein Step Twelve also includes: parsing the topological relationship data of the dynamic evacuation road network, and extracting the connection node coordinates of the diversion path and the main evacuation channel; deploying LoRa wireless beacons within a range of 20 m - 50 m from the connection nodes, with the beacon broadcast frequency of 1 time / second - 3 times / second, and the broadcast content including the node code and the path direction angle; after the application terminal receives the LoRa beacon data, compare it with the dynamic evacuation road network vector map stored locally: if the distance between the current position of the application terminal and the connection node is 15 m - 30 m and the direction angle deviation is 20 degrees - 40 degrees, trigger a voice turning prompt, if the moving trajectory of the application terminal deviates from the dynamic evacuation road network planned path for 10 seconds - 20 seconds continuously, start the vibration alarm and retransmit the dynamic evacuation road network vector map generated and updated in real time by the cloud platform data processing center to this application terminal; the voice prompt content is generated by using the TTS engine to convert the path instructions, including the turning angle value and a countdown within the range of 15 seconds - 30 seconds.

9. The intelligent control method for safety risks in chemical industrial parks according to claim 8, wherein Step Twelve also includes: when the voice turning prompt is triggered, synchronously activate the gyroscope and accelerometer of the application terminal, with a sampling frequency of 50Hz - 100Hz; calculate the angular velocity value of the turning action of the application terminal holder in real time, and set the effective turning threshold to 15 degrees / second - 30 degrees / second; If no effective turning action is detected within 3 seconds - 8 seconds after the countdown starts, extend the remaining countdown by 5 seconds - 10 seconds; if the cumulative extension times reach 2 times - 4 times, turn off the turning voice prompt and start the vibration alarm; The vibration alarm mode is adjusted according to the type of protective equipment bound to the application terminal: when the protective equipment is a heavy chemical protective suit, use 3 to 5 long vibrations lasting 1 second - 3 seconds; when the protective equipment is a light protective suit, use 8 to 12 short vibrations lasting 0.3 seconds - 0.6 seconds; the vibration intensity linearly increases with the ambient noise decibel value, with a reference intensity of 1.5G - 2.5G, and the intensity increases by 0.2G - 0.5G for every 10 decibels - 15 decibels increase in noise.

10. The intelligent control method for safety risks in chemical industrial parks according to claim 9, wherein, Step Twelve also includes: real-time collect three-axis acceleration data through the accelerometer of the application terminal, and keep the sampling frequency at 50Hz - 100Hz; when the standard deviation of the Z-axis acceleration exceeds 0.5G - 8G within 3 seconds - 5 seconds continuously, determine that the application terminal is in a violent motion state; In the violent motion state, the vibration alarm and the LED indication module are synchronously activated: when the protective equipment is a heavy chemical protective suit, the red light pulse frequency is increased to 3Hz - 5Hz, and the vibration intensity is increased by 0.5G - 1.0G; when the protective equipment is a light protective suit, the green light pulse frequency is increased to 5Hz - 8Hz, and the vibration mode is switched to intermittent strong vibration, with a working cycle of vibrating once every 0.2 seconds to 0.5 seconds, and the interval between each vibration is 0.1 seconds to 0.3 seconds; The vibration intensity compensation value is dynamically adjusted according to the motion intensity, including: the reference intensity is the set value in the protective equipment adaptation stage; for every 0.2G - 0.4G increase in the standard deviation of the Z-axis acceleration, the vibration intensity is increased by 0.1G - 0.3G; The brightness of the light pulse increases with the smoke concentration. For every 100mg / m³ - 200mg / m³ increase in the smoke concentration, the LED output lumen value is increased by 30 lumens - 80 lumens.

Images