Urban open fire area emergency supervision system and method based on internet of things large model

Abstract

The application provides a city open fire area emergency supervision system and method based on an Internet of Things large model, and relates to the field of gas safety.The method is executed by an emergency supervision management platform, and comprises the following steps: determining a first dangerous area based on the output flow rate of at least one regional gas storage tank and the gas use flow rate of at least one gas equipment, and generating a unmanned aerial vehicle collection instruction; based on the unmanned aerial vehicle collection instruction, controlling the unmanned aerial vehicle to travel to the first dangerous area and collecting an equipment image sequence; based on the equipment image sequence, determining a flame feature through a feature recognition large model; based on the flame feature and the air flow rate, determining the risk of explosion of the first dangerous area; based on the risk of explosion, determining a valve closing instruction, and closing the target valve according to the valve closing instruction.The application realizes accurate monitoring, dynamic risk assessment and automatic emergency control through multi-source data fusion and intelligent analysis, and improves the response speed and reliability of gas safety management.

Description

Technical Field

[0001] This manual relates to the field of gas safety, and in particular to an emergency monitoring system and method for urban open flame areas based on a large-scale Internet of Things (IoT) model. Background Technology

[0002] With the advancement of smart city construction, the demand for gas safety emergency management is becoming increasingly urgent. As IoT technology develops, the monitoring of gas usage scenarios is gradually shifting towards intelligence and automation. However, traditional gas monitoring methods are insufficient in distinguishing between normal open flame use and potential deflagration hazards, easily leading to misjudgments and failing to effectively address complex gas safety risks.

[0003] Therefore, it is necessary to provide an emergency monitoring system and method for urban open flame areas based on a large IoT model, in order to accurately identify the characteristics of open flame use, monitoring personnel, and gas filling, and to remotely control relevant equipment in a timely manner to reduce the risk of gas explosion. Summary of the Invention

[0004] To address the challenges of accurately identifying open flame usage scenarios, distinguishing between normal fire use and potential explosion hazards, and enabling remote emergency control in a smart city environment, this manual provides an urban open flame area emergency monitoring system and method based on an IoT big data model.

[0005] The invention includes an emergency monitoring system for urban open flame areas based on an Internet of Things (IoT) big data model. The system includes an emergency monitoring management platform configured to execute the following urban open flame area management method based on an IoT big data model.

[0006] The invention includes an emergency monitoring method for urban open flame areas based on an Internet of Things (IoT) big data model. The method is executed by an emergency monitoring and management platform and includes: determining a first hazardous area based on the output flow rate of a gas storage tank in at least one area and the gas usage flow rate of at least one gas appliance; generating a drone acquisition command based on the first hazardous area; controlling the drone to travel to the first hazardous area and acquire a sequence of device images based on the drone acquisition command; determining flame characteristics based on the device image sequence using a feature recognition big data model (a machine learning model); determining the explosion risk of the first hazardous area based on the flame characteristics and airflow rate; and determining a valve closing command based on the explosion risk and closing a target valve according to the valve closing command.

[0007] The beneficial effects of the above invention include, but are not limited to: (1) By integrating multi-dimensional data such as the output flow rate of the gas storage tank, the flow rate of the gas equipment, and the flame characteristics collected by the drone, and combining them with machine learning model analysis, the accuracy of identifying potential fire and explosion hazards in open flame areas is significantly improved, and normal fire use is effectively distinguished from dangerous scenarios; (2) Based on the dynamic parameters such as the air flow rate and flame characteristics collected in real time, the fire and explosion risk is intelligently assessed and valve closing instructions are automatically generated, realizing closed-loop management from risk warning to equipment control, and significantly shortening the emergency response time; (3) By utilizing the mobility of drones and image recognition technology, traditional manual inspections are replaced, and remote and automated monitoring of high-risk areas is realized, reducing personnel safety risks and improving regulatory efficiency. Attached Figure Description

[0008] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0009] Figure 1 This is a platform structure diagram of an urban open flame area emergency monitoring system based on an Internet of Things (IoT) big data model, as shown in some embodiments of this specification.

[0010] Figure 2 This is an exemplary flowchart of an emergency monitoring method for urban open flame areas based on an Internet of Things (IoT) big data model, as shown in some embodiments of this specification.

[0011] Figure 3 These are exemplary schematic diagrams illustrating the determination of fire and explosion risks according to some embodiments of this specification;

[0012] Figure 4 This is an exemplary flowchart of a target purging apparatus according to some embodiments of this specification. Detailed Implementation

[0013] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0014] Figure 1 This is a platform structure diagram of an urban open flame area emergency monitoring system based on an Internet of Things (IoT) big data model, as shown in some embodiments of this specification.

[0015] In some embodiments, such as Figure 1As shown, the urban open flame area emergency monitoring system 100 based on the Internet of Things large model includes an emergency monitoring user platform 110, an emergency monitoring service platform 120, an emergency monitoring management platform 130, an emergency monitoring sensor network platform 140, and an emergency monitoring object platform 150.

[0016] The emergency monitoring user platform 110 refers to a platform used by higher-level departments to comprehensively coordinate emergency monitoring. In some embodiments, the emergency monitoring user platform may include at least one user interaction device, such as a mobile phone or computer.

[0017] The Emergency Monitoring Service Platform 120 refers to an interactive service platform for receiving and transmitting data.

[0018] In some embodiments, the emergency monitoring service platform can be configured as a server, capable of interacting with the emergency monitoring user platform upstream and with the emergency monitoring management platform downstream.

[0019] The Emergency Supervision and Management Platform 130 refers to a comprehensive platform for processing and managing emergency supervision data.

[0020] In some embodiments, the emergency monitoring and management platform may be configured in a processor and / or server. The processor and / or server may process data and / or information acquired from other platforms. Based on this data, information, and / or processing results, the processor and / or server may execute program instructions to perform one or more functions described in this application.

[0021] In some embodiments, the emergency monitoring and management platform includes an emergency monitoring sub-platform and a data center.

[0022] In some embodiments, the emergency supervision sub-platform includes at least one of the following: emergency prevention sub-platform, emergency monitoring sub-platform, risk prevention sub-platform, and emergency response sub-platform.

[0023] An emergency prevention sub-platform refers to a management platform for assessing and preventing emergency events. For example, an emergency prevention sub-platform can be configured for natural disaster prevention, accident prevention, public health emergency prevention, social security incident prevention, and other scenarios requiring emergency prevention.

[0024] An emergency monitoring sub-platform refers to a platform used for monitoring, collecting, and analyzing emergency event data. For example, an emergency monitoring sub-platform can be configured for monitoring natural disasters, accidents, public health emergencies, social security incidents, and other scenarios requiring emergency monitoring.

[0025] A risk prevention sub-platform is a platform used to identify potential risks, assess risk levels, and implement risk reduction strategies. For example, a risk prevention sub-platform can be configured for natural disaster prevention, accident disaster prevention, public health emergency prevention, social security incident prevention, and other scenarios requiring risk prevention.

[0026] An emergency response sub-platform is a platform used to coordinate, dispatch, and execute emergency plans after an emergency occurs. For example, an emergency response sub-platform can be configured for natural disaster response, accident response, public health emergency response, social security incident response, and other scenarios requiring emergency response.

[0027] In some embodiments, a data center includes a database, a data processing model library, and computing units.

[0028] Databases are used to collect, store, and manage large amounts of data related to emergency management. Examples include MySQL, PostgreSQL, InfluxDB, and Prometheus.

[0029] A data processing model library refers to a collection of data processing models used for emergency management data processing.

[0030] A computing unit is a functional module used to perform arithmetic, logical, and other instruction operations. Computing units may include, but are not limited to, central processing units (CPUs).

[0031] In some embodiments, the emergency monitoring management platform interacts upward with the emergency monitoring service platform and downward with the emergency monitoring sensor network platform.

[0032] In some embodiments, the emergency monitoring and management platform is configured to: determine a first hazardous area based on the output flow rate of a gas storage tank in at least one area and the gas usage flow rate of at least one gas device; generate a drone acquisition command based on the first hazardous area; control the drone to travel to the first hazardous area and acquire a sequence of device images based on the drone acquisition command; determine flame characteristics based on the device image sequence using a feature recognition model; the feature recognition model is a machine learning model; determine the explosion risk of the first hazardous area based on the flame characteristics and air flow rate; determine a valve closing command based on the explosion risk, and close the target valve according to the valve closing command.

[0033] The emergency monitoring sensor network platform 140 refers to a management platform that transmits emergency monitoring-related sensor data or information. In some embodiments, the emergency monitoring sensor network platform may include communication equipment, servers, and various gateway devices.

[0034] In some embodiments, the emergency monitoring sensor network platform interacts upward with the emergency monitoring management platform and downward with the emergency monitoring object platform.

[0035] The emergency monitoring object platform 150 refers to a platform for collecting emergency monitoring data and implementing commands. In some embodiments, the emergency monitoring object platform 150 may include various monitoring, sensing, and interactive devices, such as gas storage tanks, drones, gas valves, pressure sensors, liquid level sensors, etc.

[0036] In some embodiments, further detailed descriptions of the IoT-based big data-driven urban open flame area emergency monitoring system 100 and its implementation method for IoT-based big data-driven urban open flame area management can be found in this specification. Figures 2-4 Related content.

[0037] It should be noted that the above description of the urban open flame area emergency monitoring system 100 and its modules based on an IoT big data model is for convenience only and should not be construed as limiting this specification to the scope of the embodiments described. It is understood that those skilled in the art, after understanding the principles of the system, may arbitrarily combine the various modules or construct subsystems connected to other modules without departing from these principles. In some embodiments, Figure 1 The emergency monitoring user platform 110, emergency monitoring service platform 120, emergency monitoring management platform 130, emergency monitoring sensor network platform 140, and emergency monitoring object platform 150 disclosed herein can be different modules within a single system, or a single module can implement the functions of two or more of the aforementioned modules. For example, the modules can share a single storage module, or each module can have its own separate storage module. Such variations are all within the scope of protection of this specification.

[0038] Figure 2 This is an exemplary flowchart illustrating an emergency monitoring method for urban open flame areas based on an IoT big data model, according to some embodiments of this specification. Figure 2 As shown, the emergency monitoring process 200 includes the following steps. In some embodiments, the emergency monitoring process 200 may be executed by an emergency monitoring management platform.

[0039] Step S210: Determine the first hazardous area based on the gas storage tank output flow rate in at least one area and the gas usage flow rate of at least one gas device.

[0040] The gas storage tank output flow rate refers to the speed at which gas flows out of the gas storage tank outlet. In some embodiments, for centralized gas supply scenarios, the gas storage tank output flow rate may include the output flow rate of the main gas control valve, etc.

[0041] Gas usage flow rate refers to the speed at which gas is used after it enters the gas equipment.

[0042] In some embodiments, the gas storage tank and gas equipment are pre-connected to the Internet of Things, and the emergency monitoring and management platform can obtain the output flow rate of the gas storage tank and the gas usage flow rate of the gas equipment through the emergency monitoring object platform.

[0043] The first danger zone refers to areas that require detailed inspection and may pose hazards such as gas leaks. Examples include the kitchens or food courts of multiple restaurants in residential areas or commercial complexes.

[0044] In some embodiments, the emergency monitoring and management platform can pre-divide the management area into multiple zones according to preset rules. Taking the management of commercial complexes as an example, the emergency monitoring and management platform can divide zones in various ways. For example, one commercial complex may correspond to one zone, or a floor within a commercial complex may correspond to one zone.

[0045] In some embodiments, the emergency monitoring and management platform can calculate the difference between the total gas tank output velocity of all gas tanks and the total gas usage velocity of all gas appliances in each region, and identify regions where the difference is greater than a first threshold as first danger zones. The first threshold can be set by technicians based on experience.

[0046] Step S220: Generate drone data collection instructions based on the first danger zone.

[0047] A drone data collection command refers to an instruction given by a drone to collect data from a target area. The target area is the region where the drone needs to collect data. In some embodiments, the drone data collection command may include the drone's flight path, the target area, and collection parameters. In some embodiments, collection parameters may include, but are not limited to, sampling altitude, sampling time, sampling frequency, and sensor type. Collection parameters can be obtained by querying a preset table based on the target area.

[0048] In some embodiments, the emergency monitoring and management platform can identify the first danger zone as the target area for the drone, and determine the drone's travel path and data collection parameters based on the first danger zone, thereby generating corresponding drone data collection instructions.

[0049] Step S230: Based on the drone acquisition command, control the drone to travel to the first danger zone and acquire the equipment image sequence.

[0050] An equipment image sequence refers to a continuous set of images of gas equipment taken in chronological order. In some embodiments, an equipment image sequence may include equipment images of gas equipment at multiple time points.

[0051] In some embodiments, the emergency monitoring and management platform can control the drone to travel to the first danger zone based on the drone's acquisition instructions, and use the camera on the drone to take pictures of the gas equipment in the first danger zone, thereby acquiring a sequence of equipment images.

[0052] Step S240: Based on the device image sequence, determine the flame characteristics through a feature recognition model.

[0053] Flame characteristics refer to parameters that characterize the presence of a flame within the first hazard area, and the state of any existing flame. In some embodiments, flame characteristics include flame height, flame burning time, etc. Flame burning time refers to the time during which a currently burning flame is continuously monitored, such as the time between the flame's initial burst and the last time point in the device's image sequence.

[0054] In some embodiments, the emergency monitoring and management platform can identify flame characteristics by recognizing equipment image sequences using a feature recognition big data model.

[0055] The feature recognition big model is a model used to extract flame features from a sequence of device image data acquired by a drone. In some embodiments, the feature recognition big model is a machine learning model. For example, the feature recognition big model may include any one or a combination of a convolutional neural network (CNN) model or other custom model architectures.

[0056] In some embodiments, the input to the feature recognition large model may include a sequence of device images, and the output may be flame features.

[0057] In some embodiments, the feature recognition large model can be obtained by training a large number of first training samples with a first label. In some embodiments, the first training samples may include a sequence of sample device images, and the first label may be a flame feature corresponding to the first training sample. In some embodiments, the first training samples may be obtained based on historical data, and the first label corresponding to the first training sample may be obtained by manual annotation.

[0058] In some embodiments, the emergency monitoring and management platform can be trained using various methods based on a first training sample and a first label. For example, it can be trained using gradient descent. As an example only, multiple first training samples with the first label can be input into an initial feature recognition model. A loss function is constructed using the first label and the results of the initial feature recognition model. The parameters of the initial feature recognition model are then iteratively updated based on the loss function. The model training is complete when the loss function of the initial feature recognition model meets preset conditions, resulting in a trained feature recognition model. These preset conditions may include loss function convergence, the number of iterations reaching a threshold, etc.

[0059] Step S250: Based on flame characteristics and air velocity, determine the explosion risk of the first hazard zone.

[0060] Explosion risk refers to the likelihood of a gas explosion.

[0061] In some embodiments, the emergency monitoring and management platform obtains flame height and flame burning time based on flame characteristics. Flame characteristics with flame height exceeding a preset threshold and burning time less than a preset time threshold are identified as target flame characteristics. After normalizing the flame height and air velocity corresponding to the target flame characteristics, the weighted value obtained from the weighted calculation is used to determine the explosion risk of the first hazard zone. The weights can be set based on experience.

[0062] Step S260: Based on the risk of combustion and explosion, determine the valve closing command, and close the target valve according to the valve closing command.

[0063] A valve closing command is an instruction that triggers a valve to close. In some embodiments, a valve closing command may include a target valve, etc.

[0064] A target valve refers to a specific valve that needs to be shut down based on the results of a gas explosion risk assessment. In some embodiments, the target valve may include valves on gas storage tanks, gas valves connected to gas equipment, etc.

[0065] In some embodiments, the emergency monitoring and management platform can determine the risk control area based on the risk of combustion and explosion, and identify the gas valves connected to all gas equipment in the risk control area, as well as the valves directly connected to the gas storage tank, as target valves.

[0066] For example, the emergency monitoring and management platform can designate areas with a risk of explosion exceeding a preset risk threshold as control areas, and designate areas directly adjacent to the control areas and the control areas as risk control areas, and generate valve closing instructions to shut down all gas valves in the risk control areas.

[0067] In some embodiments, the emergency monitoring and management platform can identify target valves based on the risk of combustion and explosion, including all gas valves connected to gas equipment and valves directly connected to gas storage tanks within the first hazardous area.

[0068] As an example, in a food court scenario within a commercial complex, multiple food vendors use liquefied petroleum gas (LPG) cylinders. When the emergency monitoring and management platform detects an abnormal difference between the output flow rate of the cylinders and the gas flow rate used by at least one gas appliance in the area, it indicates a potential gas leak risk. The platform immediately designates this area as the first danger zone. Since at least one gas appliance is still operating, the leaked gas could potentially ignite and explode upon contact with an open flame. In this situation, the platform generates a drone acquisition command and controls the drone to travel to the first danger zone. It then captures a sequence of images of the equipment in real time and inputs them into a feature recognition model for flame characteristic analysis. If flame characteristics are identified, the platform dynamically calculates the explosion risk of the first danger zone based on airflow velocity and simultaneously generates a valve closing command to remotely shut off the target valve to block the gas source.

[0069] In some embodiments of this specification, IoT technology is used to collect real-time data on the output flow rate of gas storage tanks and the flow rate of gas equipment, enabling rapid identification of potential primary hazard areas and early warning of risks such as gas leaks. Drones are quickly dispatched to hazardous areas to acquire equipment image sequences, avoiding the delays and risks associated with manual inspections. Simultaneously, large-scale feature recognition models are used to analyze the equipment image sequences, accurately identifying flame characteristics and preventing misjudgments or omissions. Valve closing commands are determined based on the risk of combustion and explosion, effectively cutting off the gas supply, reducing the likelihood of combustion and explosion or mitigating the hazards caused by combustion and explosion, and ensuring the safety of personnel and property.

[0070] Figure 3 This is an exemplary schematic diagram illustrating the determination of fire and explosion risks according to some embodiments of this specification.

[0071] In some embodiments, the emergency monitoring and management platform is further configured to: perform grid processing on the first hazardous area to obtain multiple grid cells; analyze the state characteristics of the target personnel based on human trajectory characteristics; for one of the multiple grid cells, determine the explosion risk of the grid cell based on the state characteristics, flame characteristics, and air velocity of the target personnel in the grid cell; designate the grid cells with explosion risks greater than the risk threshold as explosion grids; and determine the explosion risk of the first hazardous area based on the explosion risks of the explosion grids.

[0072] A grid cell refers to a regional unit obtained by further dividing the first danger zone.

[0073] In some embodiments, the size of the grid cells can be preset based on experience.

[0074] In some embodiments, the size of the grid cell is directly related to the number of conscious individuals within the first danger zone and their even distribution at distance. For example, the emergency monitoring and management platform can determine the size of the grid cell by querying a first preset table based on the number of conscious individuals and their even distribution at distance within the first danger zone. The first preset table includes multiple sets of correspondences between the number of conscious individuals, their even distribution at distance, and the size of the grid cell within the first danger zone. The first preset table can be pre-constructed by technicians based on historical data.

[0075] A conscious individual refers to a human whose corresponding human trajectory characteristics meet preset characteristic conditions.

[0076] Human trajectory features refer to the relevant parameter characteristics used to assess the movement of people within a primary risk area. In some embodiments, human trajectory features include the movement trajectory and speed of people. In some embodiments, human trajectory features can be acquired through drones.

[0077] In some embodiments, the preset characteristic conditions may include: the fluctuation of the motion speed is less than a second threshold and / or the frequency of abrupt changes in the motion trajectory is less than a second threshold. The second threshold may be preset by a technician according to requirements. Here, the fluctuation of the motion speed refers to the difference in speed between two adjacent time points, and an abrupt change in the motion trajectory means that the angle between the motion direction between the current time point and the previous time point is greater than a preset angle value, in which case an abrupt change is considered to have occurred at the current time point.

[0078] In some embodiments, the emergency monitoring and management platform can determine whether human trajectory characteristics meet preset characteristic conditions based on the analysis of data collected by drones, thereby determining the number of conscious individuals.

[0079] In some embodiments, the emergency monitoring and management platform can characterize the conscious state of a target person based on state characteristics. For example, when the target person is conscious, the state characteristic is 0; when the target person is unconscious, the state characteristic is 1. An unconscious state can include the target person being in a coma, intoxicated, or otherwise unable to effectively perform their risk monitoring duties.

[0080] The target personnel refers to everyone in the first danger zone.

[0081] In some embodiments, the emergency monitoring and management platform can obtain the distance between any two conscious individuals within the first danger zone, calculate the average of all distances, and then divide this average by the length of the first danger zone to obtain the uniformity of the distance distribution among conscious individuals. The length of the first danger zone can be the length of the line segment between the two farthest points within the first danger zone.

[0082] In some embodiments, if the number of conscious individuals in the first danger zone is large and the distance distribution is relatively uniform, it indicates that the state in the corresponding space is highly controllable, and the size of the grid cell can be larger to improve processing efficiency; conversely, the size of the grid cell can be designed to be smaller to refine the space and more accurately determine the actual situation.

[0083] In some embodiments, the emergency monitoring and management platform can determine the explosion risk of a grid cell based on the status characteristics of target personnel, flame characteristics, and air velocity, using a first preset algorithm. Air velocity may include average wind speed, maximum wind speed, rate of change of wind speed, wind direction, turbulence intensity, etc. In some embodiments, air velocity can be obtained through weather station data, sensors mounted on drones, etc. Further explanation regarding flame characteristics and explosion risk can be found in [link to relevant documentation]. Figure 2 And its related descriptions.

[0084] For example, the first preset algorithm may be a fire and explosion risk prediction model for grid cells. In some embodiments, the fire and explosion risk prediction model may be a machine learning model. For example, the fire and explosion risk prediction model may include any one or a combination of graph neural network (GNN) models or other custom model structures.

[0085] In some embodiments, the input to the combustion and explosion risk prediction model may include a first grid graph, and the output may be the combustion and explosion risk of each node in the first grid graph. The first grid graph may consist of at least one node and at least one edge.

[0086] In some embodiments, a node corresponds to a grid cell within a first danger zone. The node attributes may include the status characteristics of the target personnel in that grid cell and flame characteristics.

[0087] In some embodiments, when the grid cells corresponding to two nodes are adjacent, the nodes are connected by an edge, the properties of which include the airflow velocity between the two grid cells.

[0088] In some embodiments, the deflagration risk prediction model can be obtained by training a large number of second training samples with second labels. In some embodiments, the second training samples may include a sample first grid map constructed based on data collected at a first historical time point, and the second label may be the deflagration risk corresponding to each node of the sample first grid map at the second historical time point. The first historical time point is before the second historical time point, and the time interval between the two can be preset.

[0089] In some embodiments, the second training sample can be obtained through historical data. The second label can be determined based on whether a combustion and explosion accident actually occurred in the grid cell corresponding to each node of the first grid map at a second historical time point. For example, if a combustion and explosion accident actually occurred, the emergency monitoring and management platform can locate the location of the combustion and explosion through monitoring images or post-accident burning conditions, and mark the combustion and explosion risk of the grid cell where the combustion and explosion location is located as 1. Alternatively, if no combustion and explosion accident occurred, the emergency monitoring and management platform can extract gas from each grid cell, detect the gas concentration, and determine the combustion and explosion risk of the grid cell by querying a second preset table, thereby determining the value of the second label. The second preset table includes multiple sets of correspondences between gas concentrations and combustion and explosion risks. The second preset table can be pre-constructed by technicians based on experimental data.

[0090] The training process of the explosion risk prediction model is similar to that of the feature recognition model; please refer to the training process of the feature recognition model.

[0091] A risk threshold is a standard value set when assessing the risk of combustion and explosion in a grid cell. Risk thresholds can be set by technicians based on experience.

[0092] In some embodiments, the risk threshold may be related to the remaining storage capacity of the gas storage tanks in the first hazardous area. The higher the remaining storage capacity of the gas storage tanks in the corresponding first hazardous area, the lower the risk threshold.

[0093] The remaining gas storage capacity refers to the amount of gas remaining in the gas storage tank at the current point in time. The remaining gas storage capacity can be obtained from the emergency monitoring platform.

[0094] In some embodiments of this specification, the emergency monitoring and management platform can more accurately assess the risk of deflagration and achieve more sensitive safety warnings by dynamically linking the risk threshold with the remaining storage capacity of the gas storage tank.

[0095] A fire and explosion grid refers to a grid cell with a high risk of fire and explosion. For example, a grid cell with a fire and explosion risk greater than a risk threshold. In some embodiments, the emergency monitoring and management platform can label fire and explosion grids numerically. For example, if a grid cell is determined to be a fire and explosion grid, it is labeled as 1; otherwise, it is labeled as 0.

[0096] In some embodiments, the emergency monitoring and management platform can determine the explosion risk of the first hazardous area based on the explosion risks of multiple explosion grids. For example, the average or maximum value of the explosion risks of multiple explosion grids can be determined as the explosion risk of the first hazardous area.

[0097] In some embodiments, the emergency monitoring and management platform can determine the explosion risk of the first hazardous area based on the weighted values ​​of multiple explosion risks of multiple explosion grids, wherein the weight of an explosion grid is related to the distribution of gas storage tanks in the explosion grid.

[0098] Gas storage tank distribution refers to the spatial location and quantity of each gas storage tank within the explosion grid.

[0099] In some embodiments, the more gas storage tanks a combustion grid corresponds to, the greater its weight.

[0100] In some embodiments of this specification, the emergency monitoring and management platform dynamically correlates the weight of the explosion grid with the distribution of its gas storage tanks, enabling the regional explosion risk assessment to more accurately reflect the actual gas risk distribution characteristics.

[0101] In some embodiments of this specification, the risk assessment is refined to each grid cell by gridding the area, avoiding the coarseness of overall area risk determination. Combining multi-dimensional data such as the target personnel's condition characteristics, flame characteristics, and air velocity makes the risk assessment more closely resemble real-world scenarios. By marking grid cells with a combustion and explosion risk greater than the risk threshold as combustion and explosion grids, high-risk areas within the first hazard area can be quickly and accurately located. Furthermore, determining the combustion and explosion risk of the first hazard area based on the combustion and explosion risk of the combustion and explosion grids provides a more comprehensive reflection of the overall combustion and explosion risk status of the first hazard area, improving the accuracy of risk assessment.

[0102] In some embodiments, the risk of combustion and explosion in the first hazardous area is also related to the risk of secondary explosion in the combustion and explosion grid; the emergency monitoring and management platform is further configured to determine the risk of secondary explosion in the combustion and explosion grid based on the flammable material coverage of the combustion and explosion grid.

[0103] Secondary explosion risk refers to the probability that after a flammable and explosive grid has flared and exploded, the surrounding grid cells may flare and explode again.

[0104] The flammable material coverage of a flammable grid refers to the ratio of the total volume of flammable materials to the total spatial volume of the flammable grid.

[0105] In some embodiments, the emergency monitoring and management platform can determine the secondary explosion risk of the explosive grid based on the flammable material coverage of the grid using a second preset algorithm. For example, the second preset algorithm can be a secondary explosion risk prediction model.

[0106] In some embodiments, the secondary explosion risk prediction model can be a machine learning model. For example, the secondary explosion risk prediction model may include any one or a combination of graph neural network (GNN) models or other custom model structures.

[0107] In some embodiments, the input to the secondary explosion risk prediction model may include a second grid graph, and the output may be the secondary explosion risk for each node of the second grid graph. The second grid graph may consist of at least one node and at least one edge.

[0108] In some embodiments, the nodes and edges of the second grid diagram are the same as those of the first grid diagram. The difference lies in that the node attributes of the nodes in the second grid diagram may include the status characteristics of the target personnel in the grid cell, flame characteristics, explosion risk, explosion grid attributes, and target coverage. Explosion grid attributes refer to whether the grid cell is an explosion grid, and target coverage refers to the coverage of flammable materials within a preset range of the grid cell.

[0109] The preset range corresponding to a grid cell is negatively correlated with the spatial openness value of the grid cell. The spatial openness value refers to a parameter that measures the spatial circulation or openness capability of a grid cell.

[0110] In some embodiments, the spatial openness value is positively correlated with the weighted sum of the distances between the grid cell and solid obstacles (such as walls, load-bearing columns, etc.), the distances between the grid cell and secondary hazards (such as oil tanks, gas tanks, power lines, electrical equipment, etc.), and the distances between the grid cell and multiple ventilation points. The weights of the distances corresponding to each ventilation point can be preset by technicians based on experience.

[0111] In some embodiments, the secondary explosion risk prediction model can be obtained by training a large number of third training samples with third labels. In some embodiments, the third training samples may include a sample third grid diagram constructed based on data collected at a first historical time point, and the third label may be the secondary explosion risk of each node in the sample third grid diagram at the third historical time point. The first historical time point is earlier than the third historical time point, and the time distance between the first historical time point and the second historical time point is less than the time distance between the first historical time point and the third historical time point.

[0112] The method for obtaining the third training sample is similar to that for obtaining the second training sample; please refer to the method for obtaining the second training sample.

[0113] The second label can be determined based on whether a combustion and explosion accident actually occurred in the grid cells corresponding to each node of the sample third grid graph at the third historical time point, and whether there is a direct edge connection between the grid cells where the combustion and explosion accident occurred and the grid cells marked as combustion and explosion grids.

[0114] For example, if a grid cell marked as a flammable / explosive grid and at least one grid cell directly connected to it via an edge (hereinafter referred to as a neighboring cell) actually experience a flammable / explosive accident, then the secondary explosion risk of that flammable / explosive grid is 1. As another example, if a grid cell marked as a flammable / explosive grid does not experience a flammable / explosive accident, or if none of the neighboring cells of a grid cell marked as a flammable / explosive grid experience a flammable / explosive accident, the emergency monitoring and management platform can extract gas from the corresponding grid cell, detect the gas concentration, and determine the average or maximum flammable / explosive risk of each neighboring cell by querying a second preset table, thereby determining the value of the second label.

[0115] The training process for the secondary explosion risk prediction model is similar to that for the combustion and explosion risk prediction model; please refer to the training process for the combustion and explosion risk prediction model.

[0116] In some embodiments, the emergency monitoring and management platform can determine a first weighted value based on the weighted values ​​of multiple explosion risks corresponding to multiple explosion grids, determine a second weighted value based on the weighted values ​​of multiple secondary explosion risks corresponding to multiple explosion grids, and determine the explosion risk of the first hazardous area based on the weighted sum of the first and second weighted values. The weights of each parameter in each weighting process can be set by technical personnel based on experience.

[0117] In some embodiments of this specification, the potential cascading hazards of hazardous areas are covered by introducing the risk of secondary explosions, ensuring a more comprehensive and accurate risk assessment. By analyzing the coverage of flammable materials within the explosion grid, high-risk areas can be accurately identified, allowing for targeted prevention and control measures to be taken in advance, reducing the probability and impact of accidents.

[0118] In some embodiments, the emergency monitoring and management platform is further configured to: determine the explosion risk of a first hazardous area based on the weighted values ​​of multiple explosion risks and multiple secondary explosion risks of multiple explosion grids, wherein the weight of the secondary explosion risk of an explosion grid is related to the number of fire-fighting equipment in the explosion grid.

[0119] Firefighting equipment refers to the equipment pre-installed within the explosive grid for firefighting purposes. Examples include fire hydrants and fire extinguishers. In some embodiments, the greater the number of firefighting equipment, the greater the handling coverage capacity of the explosive grid. Handling coverage capacity is a parameter that measures the ability to resolve firefighting problems within the explosive grid.

[0120] In some embodiments, the weight of the secondary explosion risk of a flammable grid is negatively correlated with the treatment coverage of that flammable grid. For example, the greater the treatment coverage of a flammable grid, the smaller the weight of its secondary explosion risk.

[0121] In some embodiments of this specification, a weighted assessment mechanism for combustion and explosion risks and secondary explosion risks is introduced, and the weight of secondary explosion risks is correlated with the number of fire-fighting equipment. This fully considers the actual distribution and response capabilities of fire-fighting resources, making the risk assessment more realistic.

[0122] Figure 4 This is an exemplary flowchart of a target-determining purging apparatus according to some embodiments of this specification. Figure 4 As shown, the target purging device determination process 400 includes the following steps. In some embodiments, the target purging device determination process 400 may be executed by an emergency monitoring and management platform.

[0123] Step S410: Based on the gas tank output flow rate and gas usage flow rate, combined with historical data, determine the cruise command.

[0124] For more information on the gas tank output flow rate and gas usage flow rate, please refer to [link / reference]. Figure 2 And its related descriptions.

[0125] Historical data refers to data from previous drone patrols. In some embodiments, historical data can be obtained from the database of an emergency monitoring and management platform.

[0126] Cruise commands refer to a set of parameterized instructions used to control an unmanned aerial vehicle (UAV) to perform cruise monitoring tasks. In some embodiments, cruise commands include hovering altitude, hovering position, and hovering time.

[0127] In some embodiments, the emergency monitoring and management platform can determine patrol instructions based on the gas tank output flow rate, gas usage flow rate, and historical data using various methods. For example, the emergency monitoring and management platform can determine patrol instructions using vector matching in a vector database based on the gas tank output flow rate, gas usage flow rate, and historical data.

[0128] The emergency monitoring and management platform can construct matching vectors for each region based on the current gas tank output flow rate in at least one region and the gas usage flow rate of at least one gas appliance in the corresponding region. The vector database contains multiple feature vectors and their corresponding labels. The feature vectors are constructed from the gas tank output flow rate of multiple historical regions and the gas usage flow rate of gas appliances within those regions. The labels corresponding to the feature vectors include the actual cruise commands used.

[0129] In some embodiments, the emergency monitoring and management platform selects the cruise command with the lowest noise content from historical data through multiple tests, and constructs the gas tank output flow rate and the gas usage flow rate of each gas device in the corresponding area as feature vectors, which are then stored in the vector database.

[0130] In some embodiments, the emergency monitoring and management platform can calculate the vector similarity between the vector to be matched and the feature vector, and use the feature vector that meets the second preset condition as the target vector, and use the label corresponding to the target vector as the patrol instruction. The second preset condition can be set according to the situation. For example, the highest vector similarity, etc.

[0131] Step S420: Send a cruise command to the drone and control the drone to hover at a certain position, hovering altitude and hovering time to collect and analyze air samples from at least one area to obtain the distribution of metabolites in at least one area.

[0132] Metabolite distribution refers to the distribution characteristics of various metabolites in the air within a region that are related to gas leaks or combustion. In some embodiments, metabolite distribution includes the composition and concentration of metabolites. In some embodiments, metabolite distribution can be obtained using gas sensors, portable gas detectors, sampling equipment mounted on drones, etc.

[0133] Metabolites refer to chemical components that are pre-specified for collection. Examples include carbon monoxide and methane.

[0134] Step S430: The area where the distribution of metabolites meets the preset distribution conditions is identified as the second danger zone.

[0135] The second danger zone refers to an area that requires monitoring in addition to the first danger zone.

[0136] In some embodiments, the emergency monitoring and management platform can designate areas where the distribution of metabolites meets preset distribution conditions as second hazardous areas. The preset distribution conditions may include a concentration of key metabolite components exceeding a sixth threshold. Key metabolites refer to metabolites produced by the interaction between biological and fuel gas. In some embodiments, key metabolites can be preset by technical personnel according to requirements.

[0137] In some embodiments, for a second hazardous area, the emergency monitoring and management platform can determine valve control commands based on metabolite distribution; and close the target valve based on the valve control commands.

[0138] Valve control commands are operational commands used to control the opening or closing of target gas valves. Target valves can be all or some of the gas valves in the second hazardous area.

[0139] In some embodiments, the emergency monitoring and management platform can determine valve control commands based on the distribution of metabolites that satisfy a first preset relationship.

[0140] The first preset relationship refers to the correspondence between metabolite distribution and valve control commands. For example, when the concentration of a key metabolite component exceeds the first trigger threshold, the emergency monitoring and management platform can target some gas valves located in key positions within the second hazard area, where the key positions can be preset; when the concentration exceeds a higher second trigger threshold, all gas valves in the second hazard area will be targeted.

[0141] In some embodiments described herein, the emergency monitoring and management platform dynamically generates valve control commands based on metabolite distribution and precisely closes target valves to achieve rapid gas cutoff in high-risk areas, effectively curbing the spread of deflagration risks.

[0142] Step S440: Based on the second danger zone, generate a supplementary data collection instruction.

[0143] Supplementary data collection instructions refer to instructions given by the UAV to conduct supplementary data collection on the target area.

[0144] In some embodiments, the emergency monitoring and management platform can identify the second danger zone as the target area for the drone, thereby generating supplementary data collection instructions.

[0145] In step S450, based on the supplementary acquisition command, the drone is controlled to travel to the second danger zone and acquire supplementary image sequences.

[0146] An image supplement sequence is a continuous set of images obtained by arranging images of gas equipment taken in the second danger zone in chronological order.

[0147] The image supplementary sequence is similar to the device image sequence. For methods on obtaining the image supplementary sequence, please refer to the methods for obtaining the device image sequence.

[0148] Step S460: Based on the image supplement sequence, determine the flame supplement features through a feature recognition large model.

[0149] For more information on large-scale feature recognition models, please refer to [link / reference]. Figure 2 And its related descriptions.

[0150] Flame supplementary features refer to the flame characteristics that characterize the second hazard zone.

[0151] Similar to flame features, supplementary flame features can be found in the section on methods for determining flame features.

[0152] Step S470: Based on flame replenishment characteristics and air velocity, determine the combustion and explosion risk of the second hazard zone.

[0153] The explosion risk in the second hazardous area is similar to that in the first hazardous area. For the method of determining the explosion risk in the second hazardous area, please refer to the method of determining the explosion risk in the first hazardous area.

[0154] Step S480: Based on the explosion risk of the first hazardous area and the explosion risk of the second hazardous area, determine the target purging device and its operating parameters.

[0155] A target purging device is a device used to remove impurities, contaminants, or residues from a target area using an airflow (such as compressed air, inert gas, etc.). In some embodiments, the target purging device may include a high-pressure nitrogen purging device, etc.

[0156] In some embodiments, the operating parameters of the target purging device include nitrogen flow rate and purging pressure.

[0157] In some embodiments, the emergency monitoring and management platform can determine the high-pressure nitrogen purging devices in the first and second hazardous areas with a fire and explosion risk of not less than 0 as target purging devices based on the fire and explosion risk of the first hazardous area and the second hazardous area.

[0158] In some embodiments, the emergency monitoring and management platform can determine the operating parameters of the target purging device by querying a third preset table. This third preset table includes multiple sets of correlations between combustion and explosion risks and nitrogen flow rate and purging pressure. The third preset table can be pre-constructed by technicians based on historical data.

[0159] In some embodiments, the purging pressure of the target purging device in the first hazardous area is also related to the secondary explosion risk corresponding to the first hazardous area.

[0160] The greater the risk of secondary explosion in the first hazardous area, the greater the purging pressure of the target purging device. The risk of secondary explosion in the first hazardous area can be determined based on the risk of secondary explosion in each grid cell within the first hazardous area. For example, the risk of secondary explosion in the first hazardous area can be the mean, maximum, or weighted value of the risk of secondary explosion in all the grid cells it contains, where the weight of each grid cell can be preset.

[0161] In some embodiments of this specification, the emergency monitoring and management platform dynamically correlates the purging pressure with the risk of secondary explosion, thereby enabling intelligent adjustment of the purging intensity, which removes residual gas while avoiding secondary disasters caused by improper pressure.

[0162] Step S490: Control the target purging device to perform purging based on the operating parameters.

[0163] In some embodiments, the emergency monitoring and management platform can control the high-pressure nitrogen purging device in the first and second hazardous areas to perform purging based on the operating parameters at the location where purging is required.

[0164] In some embodiments of this specification, by combining the output flow rate of the gas storage tank with the gas usage flow rate and referencing historical data, cruise commands suitable for the current environmental conditions can be intelligently generated. This guides the UAV to precisely hover and collect air samples at key locations, improving the accuracy and timeliness of hazardous area identification. By setting parameters such as hovering altitude, hovering position, and hovering time, it ensures that the UAV can conduct stable sampling for extended periods at locations most likely to experience gas leaks or diffusion, thereby obtaining more representative air samples. Furthermore, supplementary collection commands are used to further identify secondary hazardous areas, ensuring the comprehensiveness of the risk assessment. Simultaneously, based on the risk assessment results, the target purging device and its operating parameters (such as nitrogen flow rate and purging pressure) are automatically determined, ensuring the efficiency and targeting of the purging operation.

[0165] It should be noted that the above descriptions of the emergency monitoring method flow 200 and the target purging device determination flow 400 are for illustrative purposes only and do not limit the scope of this specification. Those skilled in the art can make various modifications and changes to the emergency monitoring method flow 200 and the target purging device determination flow 400 under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0166] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0167] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. An emergency monitoring system for urban open flame areas based on an Internet of Things (IoT) big data model, characterized in that: The system includes an emergency monitoring and management platform, which is configured as follows: The first hazardous area is determined based on the gas storage tank output flow rate in at least one area and the gas usage flow rate of at least one gas device. Generate drone data collection instructions based on the first dangerous area; Based on the drone acquisition command, the drone is controlled to travel to the first danger zone and acquire a sequence of equipment images; Based on the device image sequence, flame characteristics are determined using a feature recognition model. The feature recognition model is a machine learning model; Based on the flame characteristics and air velocity, the risk of combustion and explosion in the first hazardous area is determined; Based on the aforementioned risk of combustion and explosion, a valve closing command is determined, and the target valve is closed according to the valve closing command.

2. The system according to claim 1, characterized in that, The emergency monitoring and management platform is further configured as follows: The first hazardous area is meshed to obtain multiple mesh cells; Analyze the state characteristics of target personnel based on human trajectory features; For one of the multiple grid cells, the explosion risk of the grid cell is determined based on the state characteristics of the target personnel in the grid cell, the flame characteristics, and the air velocity. The grid cells with a risk of combustion and explosion greater than the risk threshold are defined as combustion and explosion grids; Based on the explosion risk of the explosion grid, the explosion risk of the first hazardous area is determined.

3. The system according to claim 2, characterized in that, The risk threshold is related to the remaining storage capacity of the gas storage tank in the first hazardous area.

4. The system according to claim 2, characterized in that, The emergency monitoring and management platform is further configured as follows: The explosion risk of the first hazardous area is determined by weighting multiple explosion risks of multiple explosion grids, wherein the weight of one explosion grid is related to the distribution of gas storage tanks of the explosion grid.

5. The system according to claim 2, characterized in that, The explosion risk in the first hazardous area is also related to the secondary explosion risk of the explosion grid; the emergency monitoring and management platform is further configured to: The secondary explosion risk of the explosive grid is determined based on the flammable material coverage of the explosive grid.

6. The system according to claim 5, characterized in that, The emergency monitoring and management platform is further configured as follows: The explosion risk of the first hazardous area is determined by weighting multiple explosion risks and secondary explosion risks of multiple explosion grids, wherein the weight of the secondary explosion risk of one explosion grid is related to the number of fire-fighting equipment in the explosion grid.

7. The system according to claim 1, characterized in that, The emergency monitoring and management platform is further configured as follows: Based on the output flow rate of the gas storage tank and the gas usage flow rate, combined with historical data, a cruise command is determined, which includes hovering height, hovering position, and hovering time. The cruise command is sent to the drone, and the drone is controlled to hover at the hovering position and at the hovering altitude for the hovering time, so as to collect and analyze air samples from the at least one area to obtain the metabolite distribution in the at least one area; The region where the distribution of the metabolites meets the preset distribution conditions is identified as the second danger zone; Based on the second danger zone, generate supplementary data collection instructions; Based on the supplementary acquisition command, the drone is controlled to travel to the second danger zone and acquire supplementary image sequences; Based on the image supplement sequence, flame supplement features are determined using the feature recognition large model; Based on the flame replenishment characteristics and the air velocity, the risk of combustion and explosion in the second hazardous area is determined; Based on the fire and explosion risks of the first hazardous area and the second hazardous area, the target purging device and its operating parameters are determined, including nitrogen flow rate and purging pressure. The target purging device is controlled to perform purging based on the operating parameters.

8. The system according to claim 7, characterized in that, The emergency monitoring and management platform is further configured as follows: For the second hazardous area, valve control commands are determined based on the metabolite distribution; Based on the valve control command, the target valve is closed.

9. The system according to claim 7, characterized in that, The purging pressure of the target purging device in the first hazardous area is also related to the risk of secondary explosions corresponding to the first hazardous area.

10. An emergency monitoring method for urban open flame areas based on an Internet of Things (IoT) big data model, characterized in that: The method is executed by the emergency monitoring and management platform and includes: The first hazardous area is determined based on the gas storage tank output flow rate in at least one area and the gas usage flow rate of at least one gas device. Generate drone data collection instructions based on the first dangerous area; Based on the drone acquisition command, the drone is controlled to travel to the first danger zone and acquire a sequence of equipment images; Based on the device image sequence, flame features are determined using a feature recognition model; the feature recognition model is a machine learning model. Based on the flame characteristics and air velocity, the risk of combustion and explosion in the first hazardous area is determined; Based on the aforementioned risk of combustion and explosion, a valve closing command is determined, and the target valve is closed according to the valve closing command.

Images