Search

CN-120410231-B - Intelligent management and control system and method for safety risk of chemical industry park

CN120410231BCN 120410231 BCN120410231 BCN 120410231BCN-120410231-B

Abstract

The invention discloses an intelligent management and control system and method for safety risks of a chemical industry park, belongs to the technical field of risk management and control, and solves the problems of insufficient risk perception of a tank farm, delay of data transmission and insufficient risk avoidance of a dynamic path. The system comprises sensing equipment, a 5G communication module and an industrial Ethernet to form a dual-mode transmission network to upload data to a cloud platform, wherein the cloud platform establishes an equipment coding index by using a time sequence database, a rule engine executes leakage judgment to trigger early warning when methane is continuously 3-5 periods and 25% LEL and the pressure drop rate is 0.5MPa/min, a risk analysis module calls three-dimensional geographic information data to generate an evacuation path with the radius of 500-2000 m through Dijkstra algorithm, an upwind area with real-time wind direction is avoided and the evacuation path is preferentially connected with a fire station, and an application terminal receives early warning instructions containing the codes of the leakage equipment and the dynamic path. The system is used for quick early warning and evacuation planning of chemical safety accidents.

Inventors

  • LIAN HONG
  • ZHANG YUE
  • Wu Guanzheng

Assignees

  • 江西省邮电规划设计院有限公司

Dates

Publication Date
20260512
Application Date
20250702

Claims (7)

  1. 1. The intelligent management and control system for the safety risk of the chemical industry park is characterized by comprising sensing layer equipment, a data transmission network, a cloud platform data processing center and an application terminal, wherein the sensing layer equipment, the data transmission network and the cloud platform data processing center are deployed in the chemical industry park; The sensing layer equipment comprises a gas concentration sensor, a temperature sensor, a pressure sensor and a flame detector which are arranged in a major hazard source tank area; the data transmission network adopts a 5G communication module and industrial Ethernet dual-mode transmission architecture, and the sensing layer equipment packages acquired data and transmits the packaged acquired data to the cloud platform data processing center; The cloud platform data processing center comprises a time sequence database, a rule engine and a risk analysis module, wherein the time sequence database stores acquired data of sensing layer equipment in a time stamp format and establishes equipment coding indexes, and the rule engine presets a leakage judging rule, namely when methane concentration data associated with the same equipment coding exceeds 25% LEL in 3-5 continuous sampling periods, and the pressure data descending rate is greater than 0.5MPa/min, leakage early warning is triggered; the risk analysis module receives the early warning signals output by the rule engine, invokes equipment coordinate data, real-time meteorological data and emergency resource distribution data in the three-dimensional geographic information system, extracts wind speed data and wind direction angle data from the real-time meteorological data, detects overlapping parts of the buffer areas by adopting a spatial analysis algorithm, combines the dynamic meteorological vector data and generates a joint path avoidance area when the overlapping area ratio exceeds 40% -70%, sets an evacuation path area based on the real-time wind direction data, and covers the air direction on the leakage points by 30-60 degrees, calculating an evacuation path which takes the leakage point as the center and has a radius of 500-2000 m by using a Dijkstra algorithm, wherein the evacuation path avoids an upwind area of a real-time wind direction and is preferentially communicated with a coordinate point of a fire station, and the calculation frequency of the evacuation path is increased to be executed once every 10-30 seconds; The application terminal receives an early warning instruction and an evacuation path vector diagram, wherein the early warning instruction and the evacuation path vector diagram are issued by a cloud platform data processing center, the early warning instruction comprises a leakage equipment code, an early warning grade and a gas diffusion simulation polygon, and the evacuation path vector diagram comprises a dynamically updated path node coordinate set and estimated passing time of each node; The risk analysis module further comprises a laser radar scanning unit, wherein real-time point cloud data of a path node area are continuously acquired after an evacuation path is generated, the scanning frequency of the laser radar scanning unit is 5-10Hz, and the scanning angle range covers 30-60 degrees on two sides of the center line of the evacuation path; The risk analysis module performs difference ratio comparison on real-time point cloud data and base map elevation data of the three-dimensional geographic information system, and triggers evacuation path re-planning when the absolute difference value of the point cloud elevation value of the newly added obstacle surface relative to the base map elevation data is greater than 50 cm; The cloud platform data processing center also comprises an instruction distribution module connected with the risk analysis module and the application terminal, wherein the instruction distribution module comprises: The terminal positioning unit acquires physical position data through a GPS positioning module arranged in the application terminal, and the positioning accuracy error is less than 3 meters; The distance classification unit calculates the linear distance between each application terminal and the leakage point, and divides the linear distance into 3 priority levels, wherein the distance between each application terminal and the leakage point is less than or equal to 500 meters and is a first priority, the distance between each application terminal and the leakage point is more than 500 meters and less than or equal to 1000 meters and is a second priority, the distance between each application terminal and the leakage point is more than 1000 meters and less than or equal to 2000 meters and is a third priority, and when the distance between each application terminal and the leakage point is more than 2000 meters, early warning is automatically released; And the transmission control unit is used for directly transmitting the evacuation path vector diagram to the first priority application terminal when the evacuation path is re-planned and triggered, transmitting the data packets to the second and third priority application terminals by adopting a UDP protocol through the 5G communication module, storing the data packets into a Redis cache queue, delaying for 2-5 seconds and transmitting the data packets through the industrial Ethernet, wherein the resource allocation proportion of the transmission control unit is dynamically adjusted according to the personnel evacuation priority strategy in the emergency plan of the chemical industry park, wherein the allocation resource proportion of the first priority application terminal is 60-80% of the total bandwidth, and the residual bandwidth is allocated according to the proportion of the second priority application terminal number to the third priority application terminal number=3:1.
  2. 2. The intelligent management and control method for the safety risk of the chemical industry park is characterized by comprising the following steps of: Step one, acquiring methane concentration data in real time through a gas concentration sensor deployed in a major dangerous source tank area; step two, acquiring tank area temperature data through a temperature sensor; step three, collecting tank area pressure data through a pressure sensor; step four, acquiring flame radiation data through a flame detector; Step five, methane concentration data, temperature data, pressure data and flame radiation data acquired in the step one to the step four are sent to a cloud platform data processing center through a 5G communication module and an industrial Ethernet dual-mode transmission architecture; step six, storing the received data in a time sequence database of a cloud platform data processing center, and establishing an equipment coding index in a time stamp format; Step seven, executing leakage judgment through a rule engine, wherein when the continuous 3-5 sampling periods of methane concentration data associated with the same equipment code exceed 25% LEL, and the pressure data descending rate is greater than 0.5MPa/min, a leakage early warning signal is generated; Step eight, responding to leakage early warning signals, calling equipment coordinate data, real-time meteorological data and emergency resource distribution data stored in a three-dimensional geographic information system, extracting wind speed data and wind direction angle data from the real-time meteorological data, wherein the wind speed detection range is 1-30 m/s, the wind direction angle detection range is 0-360 degrees, performing cubic spline interpolation calculation on the wind speed data and the wind direction angle data by taking 15-40 seconds as interpolation periods to generate a wind field model of a continuous time sequence, associating wind field model output with leakage point equipment codes, storing the wind field model output as dynamic meteorological vector data, retrieving all equipment codes associated with current leakage early warning signals, extracting the coordinate data of each leakage point and the dynamic meteorological vector data, establishing a buffer area with the radius of 30-60 m by taking the coordinate of the leakage point as the center, detecting the overlapping part of the buffer area by adopting a spatial overlapping analysis algorithm, and combining the dynamic meteorological vector data and generating a joint path area when the overlapping area accounts for more than 40-70 percent; Setting an evacuation path avoidance area based on real-time wind direction data, wherein the evacuation path area covers a fan-shaped area with the wind direction of 30-60 degrees on a leakage point and is 200-800 meters away from the leakage point; step ten, calculating an evacuation path which takes a leakage point as a center and has a radius of 500-2000 meters through a Dijkstra algorithm, wherein the evacuation path bypasses an evacuation path evasion region set in the step nine and is preferentially connected to the nearest fire station coordinate point, and the calculation frequency of the evacuation path is increased to be executed once every 10-30 seconds; Eleventh, packaging the leakage equipment code, the early warning level and the gas diffusion simulation polygon into an early warning instruction; step twelve, encapsulating the evacuation path node coordinate set, the estimated transit time of each node and the real-time wind direction vector into an evacuation path vector diagram; and thirteenth step, issuing an early warning instruction and an evacuation path vector diagram to the application terminal through the 5G communication module.
  3. 3. The intelligent control method for the safety risk of the chemical industry park according to claim 2, wherein, The setting range of the path avoidance area in the step nine is expanded to be a sector area with the angle of 30-60 degrees when a single leakage point is used, the setting range of the path avoidance area is adjusted to be a composite sector area with the angle of 50-120 degrees when the path avoidance area is combined, and the distance is the range of 200-800 meters at the maximum in each leakage point; And step ten, calculating an evacuation path by adopting a hierarchical planning strategy, namely planning a main evacuation channel, connecting safety exit coordinate points with the radius of 500-2000 meters, and generating branch paths to the coding areas of all the devices based on the main evacuation channel.
  4. 4. The intelligent control method for safety risk of chemical industry park according to claim 3, wherein the branch path generation stage in the step ten comprises the following steps: S1, acquiring real-time video monitoring data along a trunk evacuation channel, and identifying the personnel density and the moving speed in the trunk evacuation channel through YOLOv target detection algorithm; S2, when the personnel density is 1-3 people/square meter and the moving speed is 0.8-1.2 meters/second, marking the main evacuation channel section as a congestion node; S3, searching a standby branch path with the radius of the congestion node within the range of 50-100 meters, and calculating the bypass distance increment and gradient data of the standby branch path; S4, selecting a standby branch path with the detour distance increment of 120-200 m and the gradient of 8-15 degrees as a shunt path; S5, topological association is carried out on the node coordinates of the diversion paths and the nodes of the trunk evacuation channels, so that a dynamic evacuation road network is generated; And the evacuation path vector diagram package of the step twelve contains topology relation data of the dynamic evacuation path network.
  5. 5. The intelligent management and control method for safety risks in a chemical industry park is characterized by further comprising the steps of analyzing topological relation data of a dynamic evacuation road network, extracting joint node coordinates of a diversion path and a trunk evacuation channel, deploying a LoRa wireless beacon in a range of 20-50 m of joint node radius, broadcasting the beacon at a frequency of 1-3 times/s, broadcasting content including node codes and path direction angles, comparing a locally stored dynamic evacuation road network vector diagram after an application terminal receives the LoRa beacon data, triggering voice steering prompt if the current position of the application terminal is 15-30 m away from the joint node and the direction angle deviation is 20-40 degrees, starting vibration warning and retransmitting the dynamic evacuation road network vector diagram which is generated and updated in real time by a cloud platform data processing center to the application terminal if the movement track of the application terminal is deviated from a dynamic evacuation road network planning path for 10-20 seconds, and generating a TTS engine conversion path instruction for voice prompt content including a steering angle value and a countdown time range of 15-30 seconds.
  6. 6. The intelligent control method for the safety risk of the chemical industry park according to claim 5, wherein the step twelve further comprises the steps of synchronously activating a gyroscope and an accelerometer of the application terminal when voice steering prompt is triggered, wherein the sampling frequency is 50-100Hz; if no effective steering action is detected within 3-8 seconds after the countdown is started, prolonging the remaining countdown for 5-10 seconds, if the accumulated prolonging times reach 2-4 times, closing a steering voice prompt and starting a vibration alarm; The vibration alarm mode is adjusted according to the type of protective equipment bound by an application terminal, wherein when the protective equipment is heavy protective clothing, long vibration lasting for 1-3 seconds is adopted for 3-5 times, when the protective equipment is light protective clothing, short vibration lasting for 0.3-0.6 seconds is adopted for 8-12 times, the vibration intensity is linearly improved along with the decibel value of environmental noise, the reference intensity is 1.5-2.5G, and the noise is improved by 0.2-0.5G every 10-15 decibels.
  7. 7. The intelligent control method for the safety risk of the chemical industry park according to claim 6, wherein the step twelve further comprises the steps of collecting three-axis acceleration data in real time through an accelerometer of the application terminal, keeping the sampling frequency at 50-100Hz, and judging that the application terminal is in a severe motion state when the standard deviation of the Z-axis acceleration exceeds 0.5-8G in 3-5 seconds; When the protective equipment is light protective clothing, the green light pulse frequency is raised to 5-8Hz, the vibration mode is switched to intermittent strong vibration, the working period is that the vibration is carried out once every 0.2 to 0.5 seconds, and the vibration interval is 0.1 to 0.3 seconds; the vibration intensity compensation value is dynamically adjusted along with the motion intensity, and comprises that the reference intensity is a set value in the adaptation stage of the protective equipment, and the vibration intensity is improved by 0.1-0.3G when the standard deviation of the Z-axis acceleration is increased by 0.2-0.4G; The brightness of the light pulse is improved along with the smoke concentration, and the output lumen value of the LED is improved by 30-80 lumens when the smoke concentration is increased by 100-200mg/m < 3 >.

Description

Intelligent management and control system and method for safety risk of chemical industry park Technical Field The invention relates to the technical field of chemical industry park risk management and control, in particular to an intelligent chemical industry park safety risk management and control system and method. Background The safety risk monitoring of important dangerous source tank areas in chemical industry parks has the problems of response lag and low evacuation efficiency, which are mainly caused by multi-dimensional technical defects. At the risk perception level, existing systems typically employ only a single type of sensor for monitoring, e.g., relying on gas concentration detection or pressure detection alone, and it is difficult to capture the risk of multi-parameter coupling. Because the early stage of gas leakage is often expressed as a composite characteristic of pressure dip and concentration ramp, a single sensor cannot establish a correlation rule among parameters, so that the early stage leakage early warning leakage rate is high. Meanwhile, the deployment positions of the traditional sensors are fixed outside the tank body, so that the response to local concentration gradient changes generated by micro leakage is slow, and the optimal treatment time is delayed. In the data transmission link, the existing system generally depends on a single network system for communication. When a wireless network is adopted, electromagnetic shielding effect generated by dense metal tanks in a chemical industry park is easy to cause signal attenuation, data packet loss frequently occurs in a high concurrency data transmission process, key parameter change trend (such as pressure instantaneous dropping rate) is difficult to upload completely, when a wired industrial bus is adopted, network expansibility is poor, newly-added monitoring points need to be rewiring, bandwidth allocation is stiff, and flow peaks under sudden risk events cannot be adapted. These defects cause timestamp dislocation or key frame deletion of data received by the cloud platform, and directly affect timeliness of subsequent risk analysis. In the risk treatment stage, the existing path planning method has obvious limitations. On the one hand, the evacuation path is generated by depending on static electronic map data, and weather monitoring information is not fused in real time. Because the chemical gas diffusion range is directly affected by wind direction and wind speed, a static path may guide personnel into a gas diffusion area, causing secondary injury. On the other hand, path planning is not dynamically associated with emergency resources, such as not preferentially connecting the nearest available fire station or medical point, and the path detours for the rescuer to reach the leak point, delaying initial disposal efficiency. In addition, the evacuation instruction issued by the traditional system lacks of spatial information refinement, only provides text early warning or a simple schematic diagram, and on-site personnel cannot quickly position the geometrical relationship between the self position and the optimal path. The problems are caused by the technical bottlenecks of three aspects, namely firstly, the real-time fusion of multi-source heterogeneous sensor data is required to solve the problem of parallel processing of high-throughput time sequence data, particularly, different sampling frequencies and dimensions exist in parameters such as gas concentration, pressure, temperature and the like, the complexity of time sequence alignment and feature extraction is required to be overcome in establishing a cross-parameter association rule, secondly, the dynamic path planning is required to break through the high-efficiency coupling calculation obstacle of meteorological data and a geographic information system, the influence of wind direction change on the path is required to respond in seconds, but the calculation cost of a traditional space analysis algorithm is high, and finally, the synchronous issuing of an early warning instruction and an evacuation path is required to balance the reliability and the real-time performance, the data volume of a high-precision vector map is large, and the technical contradiction exists in realizing low-time delay transmission under a limited bandwidth. These difficulties have long limited the improvement in safety control performance of chemical parks. Disclosure of Invention The invention provides an intelligent management and control system and method for safety risks of a chemical industry park, and aims to solve the problems of insufficient multi-parameter risk perception, insufficient data transmission instantaneity and unfused real-time weather and emergency resources of a dynamic evacuation path of a major hazard source tank area of the chemical industry park. The existing system has the problems of high leakage early warning and leakage report rate, evacuation path crossin