Railway rescue vehicle disaster identification method and device

Abstract

The application provides a track rescue vehicle disaster identification method and device. The method comprises the following steps: acquiring real-time sensing data collected by a sensing device arranged on a track rescue vehicle; obtaining a lateral offset distance of a lowest temperature point in a thermal imaging picture relative to a picture center line based on thermal imaging data; storing a longitudinal mileage, a gas type, a gas concentration, weather data and the lateral offset distance as a queue element in a data queue; acquiring a longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and the longitudinal mileage in the data queue; acquiring a plurality of target queue elements that are most recently acquired in the data queue in the case that a plurality of continuous longitudinal concentration gradients are positive; acquiring a longitudinal mileage of a combustion and explosion early warning zone based on the longitudinal mileage, the gas type, the gas concentration and the weather data in the plurality of target queue elements; and obtaining a disaster identification result in combination with the lateral offset distance, thereby improving the universality and applicability of disaster identification.

Description

Technical Field

[0001] This application relates to the field of tunnel disaster identification technology, and in particular to a method and device for identifying disasters using a rail rescue vehicle. Background Technology

[0002] When a disaster occurs in an urban rail transit tunnel section, resulting in a gas leak or a hazardous chemical leak, the current common approach is to send rescue personnel with portable gas detectors to enter the tunnel on foot or to use a mobile platform for reconnaissance. Both methods have obvious limitations. Not only do they expose personnel to harmful environments and pose a high risk, but they also face the following long-standing unresolved problems in practical applications.

[0003] First, the algorithms for locating leak sources rely too heavily on pre-built maps. Currently, there are methods that collect multi-source data through mobile platforms, combine this with SLAM algorithms to estimate poses, and match the results with 3D panoramic images to trace leak sources. However, these methods are difficult to use directly in unfamiliar tunnels or scenarios without prior maps.

[0004] Second, different gases exhibit significantly different diffusion behaviors in tunnel environments. Light gases such as methane, which are lighter than air, and heavy gases such as hydrogen sulfide, which are heavier than air, have completely different concentration decay patterns and hazardous area morphologies. However, existing vehicle-mounted gas prediction and monitoring systems rarely distinguish between gas types, often employing a uniform diffusion model or a fixed safety distance, leading to significant discrepancies between prediction results and actual conditions. Summary of the Invention

[0005] This application provides a method and apparatus for disaster identification using a rail rescue vehicle, aiming to at least partially address one of the technical problems in related technologies. The technical solution of this application is as follows:

[0006] In a first aspect, embodiments of this application propose a method for disaster identification using a rail rescue vehicle, including:

[0007] The real-time sensing data collected by the sensing devices deployed on the track rescue vehicle includes longitudinal mileage, gas type, gas concentration, meteorological data, and thermal imaging data.

[0008] Based on the thermal imaging data, the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image is obtained, and the longitudinal distance, gas type, gas concentration, meteorological data and the lateral offset distance are stored as a queue element in the data queue.

[0009] Based on the gas concentration and longitudinal mileage in the data queue, the longitudinal concentration gradient of adjacent queue elements in the data queue is obtained, and if multiple consecutive longitudinal concentration gradients are positive, the most recently obtained target queue elements in the data queue are obtained.

[0010] Based on the longitudinal mileage, gas type, gas concentration and meteorological data in the multiple target queue elements, the longitudinal mileage of the explosion warning zone is obtained, and based on the longitudinal mileage of the explosion warning zone and the lateral offset distance in the multiple target queue elements, the disaster identification result is obtained.

[0011] Based on the disaster identification results, an early warning is issued.

[0012] Secondly, embodiments of this application propose a disaster identification device for a rail rescue vehicle, comprising:

[0013] The data acquisition module is used to acquire real-time sensing data collected by the sensing devices deployed on the rail rescue vehicle. The real-time sensing data includes longitudinal mileage, gas type, gas concentration, meteorological data, and thermal imaging data.

[0014] The data processing module is used to obtain the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image based on the thermal imaging data, and store the longitudinal distance, gas type, gas concentration, meteorological data and the lateral offset distance as a queue element into the data queue.

[0015] The gradient processing module is used to obtain the longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and longitudinal mileage in the data queue, and to obtain the most recently acquired target queue elements in the data queue when multiple consecutive longitudinal concentration gradients are positive.

[0016] The disaster identification module is used to obtain the longitudinal mileage of the explosion warning zone based on the longitudinal mileage, gas type, gas concentration and meteorological data in the multiple target queue elements, and to obtain the disaster identification result based on the longitudinal mileage of the explosion warning zone and the lateral offset distance in the multiple target queue elements.

[0017] The early warning output module is used to output early warnings based on the disaster identification results.

[0018] This application has the following advantages and beneficial effects:

[0019] This solution overcomes the shortcomings of existing leak source localization algorithms. By solving the concentration gradient, this method can quickly determine the explosion warning zone using data such as longitudinal mileage and gas concentration sampled during the movement of the rescue vehicle along the track, without pre-mapped or fixed sensor networks. The method has strong applicability and can solve the problem that existing solutions are not suitable for unfamiliar tunnels or scenarios without prior maps.

[0020] This disaster identification method uses thermal imaging data to obtain the lateral offset distance of the leak source relative to the centerline of the track. Combined with the explosion warning zone, it can identify the location of the leak point, thereby determining the high-risk area and safety boundary of the accident.

[0021] This solution uses appropriate diffusion models to predict the concentration distribution of gases with different physical properties, such as methane and hydrogen sulfide, and delineates explosion warning zones accordingly, thus adapting to the hazardous characteristics of different leaked media.

[0022] This solution considers both gas concentration conditions and ignition source conditions to improve accuracy and avoid missing potential high-risk scenarios.

[0023] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0024] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0025] Figure 1 This is a flowchart illustrating the disaster identification method for rail rescue vehicles provided in the embodiments of this application;

[0026] Figure 2 This is a schematic diagram of the structure of the rail rescue vehicle provided in the embodiments of this application;

[0027] Figure 3 This is a schematic diagram of the structure of the track rescue vehicle disaster identification device provided in the embodiments of this application.

[0028] Explanation of reference numerals in the attached figures: 1 is the vehicle-mounted display control panel, 2 is the ultrasonic mini weather station, and 3 is the disaster perception module. Detailed Implementation

[0029] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0030] The following description, with reference to the accompanying drawings, illustrates a method and apparatus for identifying disasters using a rail rescue vehicle, as described in this application.

[0031] Figure 1 This is a flowchart illustrating a disaster identification method for a rail rescue vehicle provided in an embodiment of this application.

[0032] It should be noted that the executing entity of the track rescue vehicle disaster identification method in this application embodiment is the track rescue vehicle disaster identification device in this application embodiment. The track rescue vehicle disaster identification device can be configured in an electronic device so that the electronic device can perform the track rescue vehicle disaster identification function.

[0033] like Figure 1 As shown, the disaster identification method for the rail rescue vehicle includes the following steps:

[0034] Step S101: Obtain real-time sensing data collected by the sensing devices deployed on the rail rescue vehicle. The real-time sensing data includes longitudinal mileage, gas type, gas concentration, meteorological data, and thermal imaging data.

[0035] In some embodiments, the sensing device includes a positioning module, a thermal imaging module, a light gas detector, a heavy gas detector, a timing module, and a miniature weather station. In this embodiment, the positioning module can obtain longitudinal mileage, the light or heavy gas detector can obtain gas concentration and gas type, the thermal imaging module can obtain thermal imaging data, and the miniature weather station can obtain meteorological data, which may include wind direction and wind speed. The timing module adds a unified timestamp to the data collected by the sensing device. As an example, such as... Figure 2 As shown, the hardware deployed on the rail rescue vehicle includes: an onboard display control panel 1, an ultrasonic miniature weather station 2, and a disaster perception module 3. The disaster perception module 3 includes: a visible light / infrared thermal imaging module with a resolution of 640×512, a temperature measurement range of -20℃ to 550℃, and an accuracy of ±2℃; a gas detection module, including at least methane (0~100% LEL) and hydrogen sulfide gas detection modules; a positioning module using wheel speed odometer + IMU fusion, with a relative positioning error ≤0.1 meters; and a time synchronization module to add a unified timestamp to all data. The ultrasonic miniature weather station 2 is a two-parameter type used to measure wind direction and wind speed.

[0036] Step S102: Based on the thermal imaging data, obtain the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image, and store the longitudinal mileage, gas type, gas concentration, meteorological data and lateral offset distance as a queue element into the data queue.

[0037] In some embodiments, obtaining the lateral offset distance of the lowest temperature point in a thermal imaging image relative to the image center line based on thermal imaging data includes: obtaining the lateral pixel offset of the lowest temperature point in a thermal imaging image relative to the image center line based on thermal imaging data; and obtaining the lateral offset distance based on the lateral pixel offset and a pixel distance conversion coefficient. As an example, in the thermal imaging image corresponding to the thermal imaging data, the pixel coordinates of the lowest temperature point are extracted, and the lateral pixel offset of that point relative to the image center line is obtained. (Left negative, right positive), the lateral offset distance can be obtained using the following formula:

[0038] (1)

[0039] in, The pixel distance conversion coefficient is obtained through pre-calibration, such as placing a heat source with a known lateral offset on a known track, with units of meters per pixel. This represents the horizontal pixel offset, in pixels. It represents the lateral offset distance from the lowest temperature point in the thermal image to the center line of the image, in meters.

[0040] It should be noted that, since the rail rescue vehicle is constrained by the rails, it can only move along the rails and cannot move freely to measure the lateral concentration distribution. This invention proposes an indirect method that utilizes the principle that a local low-temperature zone will form around the leak point due to the Joule-Thomson effect (i.e., throttling expansion and heat absorption) when high-pressure gas leaks. This low-temperature zone is inferred to be the location of the corresponding gas leak source. This invention can clearly capture this low-temperature zone through infrared thermal imaging, thereby indirectly determining the lateral offset distance of the leak source.

[0041] As one implementation, the data queue is a circular queue of length n, forming a sliding window. The data queue is updated in real time as the track rescue vehicle moves, discarding the oldest data and adding the newest data. For example, if the track rescue vehicle travels along the longitudinal direction (z-axis) of the tunnel (approximately 1.4 meters at a speed of 5 km / h), and n is 60 (corresponding to 60 seconds of historical data), the i-th element in the data queue is:

[0042]

[0043] in, For the data queue One element, This represents the vertical mileage in the i-th element of the data queue, in meters. For the data queue The gas concentration of each element is expressed in %LEL or ppm. For the data queue The gas types in the element include light gases that are lighter than air and heavy gases that are heavier than air; For the data queue The lateral offset distance of the lowest temperature point in the thermal image of each element relative to the center line of the image, in meters; This is the timestamp of the i-th element in the data queue, in seconds. For the data queue The wind direction in each element includes positive (wind direction is the same as the diffusion direction) and negative (wind direction is opposite to the diffusion direction); For the data queue The wind speed in each element is expressed in m / s.

[0044] It should be noted that the unit of gas concentration for different types of gases corresponds to the gas detector used. In this embodiment, the gas concentration collected by the detector used to detect methane gas concentration is %LEL, and the gas concentration collected by the detector used to detect hydrogen sulfide gas concentration is ppm.

[0045] Step S103: Based on the gas concentration and longitudinal mileage in the data queue, obtain the longitudinal concentration gradient of adjacent queue elements in the data queue, and if multiple consecutive longitudinal concentration gradients are positive, obtain the multiple most recently obtained target queue elements in the data queue.

[0046] In some embodiments, the method for obtaining the longitudinal concentration gradient of adjacent queue elements in a data queue based on the gas concentration and longitudinal mileage in the data queue includes: obtaining the concentration difference of each adjacent queue element based on the gas concentration in the data queue; obtaining the mileage difference of each adjacent queue element based on the longitudinal mileage in the data queue; and obtaining the longitudinal concentration gradient of adjacent queue elements based on the concentration difference and mileage difference of each adjacent queue element.

[0047] As an example, for each queue element in the data queue, the following formula is used to calculate the... Vertical concentration gradient of adjacent queue elements :

[0048] (2)

[0049] in, For the first The longitudinal concentration gradient of adjacent queue elements, in units of %LEL / m³ or ppm / m³. For the data queue The gas concentration of each element is expressed in %LEL or ppm. For the data queue The gas concentration of each element is expressed in %LEL or ppm. For the data queue The vertical mileage in each element, in meters; For the data queue The vertical mileage in each element is expressed in meters.

[0050] As one implementation method, when three consecutive longitudinal concentration gradients are positive, the most recently acquired target queue elements in the data queue are obtained. This embodiment of the invention considers that actual gas concentration data within tunnels may fluctuate, and directly using a single-point longitudinal concentration gradient is prone to misjudgment. Therefore, a "three consecutive points" rule is adopted: if a certain time point exists... , making , , If so, the leak source is determined to be located in front of the vehicle in the direction of travel.

[0051] Step S104: Based on the longitudinal mileage, gas type, gas concentration and meteorological data in multiple target queue elements, obtain the longitudinal mileage of the explosion warning area, and obtain the disaster identification result based on the longitudinal mileage of the explosion warning area and the lateral offset distance in multiple target queue elements.

[0052] In some embodiments, the method for obtaining the longitudinal mileage of a fire and explosion warning zone based on longitudinal mileage, gas type, gas concentration, and meteorological data in multiple target queue elements includes: obtaining the fire and explosion warning point concentration and deflagration critical point concentration corresponding to the gas type in the multiple target queue elements; obtaining the average concentration gradient based on the longitudinal mileage and gas concentration in the multiple target queue elements; obtaining the longitudinal mileage of the starting point of the fire and explosion warning zone based on the fire and explosion warning point concentration, the average concentration gradient, and the longitudinal mileage and gas concentration of the nearest queue element in the multiple target queue elements; obtaining the longitudinal distance from the deflagration critical point concentration to the fire and explosion warning point concentration based on the distance decay formula corresponding to the gas type, wind speed data in the meteorological data, the fire and explosion warning point concentration, and the deflagration critical point concentration; and obtaining the longitudinal mileage of the ending point of the fire and explosion warning zone based on the longitudinal distance and the longitudinal mileage of the starting point of the fire and explosion warning zone.

[0053] One implementation approach is to take the starting element of the queue of the most recent positive gradient segment. The gas concentration of the element at that point and longitudinal mileage For reference, let's assume the concentration at the explosion warning point is... (If 30% LEL is taken, it is the optimal engineering value determined by the upper limit of the high-precision measurement range of the sensor, the alarm threshold level of the national standard, and the principle of extrapolation conservatism. It aims to ensure the quality of positioning data, achieve early warning, and control extrapolation error.) Then the longitudinal mileage of the explosion hazard warning point (i.e., the longitudinal mileage of the starting point of the explosion warning zone) is as follows. It can be estimated using the following formula:

[0054] (3)

[0055] in, This is the average concentration gradient over the positive gradient segment, calculated using least squares linear regression. If the most recent continuous data points are taken from multiple target cohorts, there are a total of... The point, the first The coordinates of the points are , using a straight line (in, (Intercept) Fit this Points, slope This equals the average concentration gradient within that interval. The formula is as follows:

[0056] (4)

[0057] In some embodiments, the method for obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance-concentration decay formula corresponding to the gas type, wind speed data in meteorological data, the concentration at the deflagration warning point, and the deflagration critical point concentration includes: when the gas type is light gas, determining the longitudinal diffusion coefficient of the distance-concentration decay formula corresponding to the gas type based on wind direction data in meteorological data; and obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance-concentration decay formula corresponding to the gas type, the longitudinal diffusion coefficient, wind speed data in meteorological data, the concentration at the deflagration warning point, and the deflagration critical point concentration.

[0058] As one implementation method, for gases lighter than air, such as methane, the tunnel is considered as a one-dimensional pipe, and the gas concentration distribution satisfies the following convection-diffusion equation (also known as the gas diffusion model):

[0059] (5)

[0060] After continuous leakage and a certain period of time, the solution can be approximated as a steady state:

[0061] (6)

[0062] In equations (5) and (6), The longitudinal diffusion coefficient is taken as 10 m² / s (empirical value) within the tunnel. The algorithm considers wind direction; if the wind direction is consistent with the diffusion direction, the diffusion coefficient... Taking 10 m² / s (empirical value), if the wind direction is opposite to the diffusion direction, the diffusion coefficient in the formula... It is 0.3 times that of the tailwind, and is taken as 3 m² / s (empirical value). Wind speed in meteorological data, in m / s; Distance from the location of the deflagration critical concentration The gas concentration at that location is expressed in g / m³.3 ; The tunnel area is perpendicular to the wind direction, in meters (m²). 2 ; The mass of pollutants released per unit time, i.e., the emission rate of the pollution source, is expressed in g / s.

[0063] It should be noted that the longitudinal diffusion coefficient This reflects the intensity of turbulence, which is affected by various factors such as tunnel wall roughness, wind speed, and cross-sectional shape, making it impossible to measure accurately in real time. This invention, based on the typical flow field characteristics of subway tunnels, defaults to... = 10m 2 / s, this value comes from statistical analysis of measured data from multiple subway tunnels.

[0064] The above formula transformation yields the formula for distance decay with concentration for light gases, as follows:

[0065] (7)

[0066] In equation (7), This is the deflagration critical point concentration (e.g., 100% LEL). To the concentration of leaked gas from Reduce to target concentration Distance, in meters; The wind speed is from meteorological data, in m / s; the wind speed and direction in the tunnel when the tunnel is stationary at the starting point are taken. If the wind direction is the same as the diffusion direction, the diffusion coefficient is... Taking 10 m² / s, if the wind direction is opposite to the diffusion direction, then the diffusion coefficient is... It is 3 m² / s.

[0067] Therefore, for gases lighter than air, such as methane, this invention can calculate the reduction of the leaked gas concentration from the deflagration critical point concentration to the deflagration warning point concentration by replacing the target concentration in formula (7) with the deflagration warning point concentration. The longitudinal distance.

[0068] For methane, this invention identifies the concentration curve that satisfies 30% LEL ≤ Longitudinal mileage ranges ≤ 100%LEL are denoted as This zone is designated as a deflagration warning zone, where combustion or explosion may occur at any point within the zone if an ignition source is present.

[0069] It should be noted that since the gas concentration at the actual leak point is far beyond the sensor's range (usually >100% LEL), the sensor cannot measure it directly. This invention utilizes the sensor's measurement data in the high-precision range (0-30% LEL) to extrapolate a concentration gradient to determine a virtual reference point (the starting point of the deflagration warning zone, 30% LEL). The methane concentration at this point is defined as 30% LEL. Starting from this point, the downstream concentration distribution is predicted using the distance-concentration decay formula, and the area with gas concentration between 30% LEL and 100% LEL is calibrated as the deflagration warning zone.

[0070] The selection of the 30% LEL threshold for methane leakage is based on the following criteria: 1. The absolute error of commercially available methane sensors in the 0–30% LEL range is ≤ ±3% LEL. Above 30% LEL, the error becomes a reading error, which is relatively large (±10%). 30% LEL is the engineering boundary for sensor accuracy zoning, while still achieving an early warning effect; 2. Gas leakage below 30% LEL has entered a well-mixed, far-source region, and the concentration distribution follows an exponential decay law. However, the diffusion behavior in the near-source region above 30% LEL is complex and difficult to describe with a simple model. Therefore, a diffusion model is adopted, which can achieve higher accuracy. In the first stage (0–30% LEL), the system uses the high-precision measured data of the sensor to fit the concentration gradient using the least squares method, and then linearly extrapolates to obtain a virtual reference point (30% LEL position). Gas leakage below 30% LEL has entered a well-mixed, far-source region, and the concentration distribution follows an exponential decay law, thus this mathematical model can achieve high accuracy. In the second stage (30%–100% LEL), the system uses a virtual reference point as a starting point and employs a diffusion model to predict the downstream concentration distribution, thereby delineating the explosion warning zone. For the diffusion of combustible methane gas, the delineation of the explosion warning zone differs under downwind and upwind conditions. This invention fully considers the influence of wind direction, making it more realistic.

[0071] Different gas diffusion models are used for different gases. Formula (5) above is applicable to the diffusion model of gases lighter than air. The following introduces the diffusion model applicable to gases heavier than air and the solution formula for the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration derived from the diffusion model.

[0072] In some embodiments, the method for obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance-concentration decay formula corresponding to the gas type, wind speed data in meteorological data, the concentration at the deflagration warning point, and the deflagration critical point concentration includes: when the gas type is heavy gas, obtaining the heavy gas characteristic length based on the tunnel width of the tunnel where the rail rescue vehicle is located and the wind speed data in meteorological data; and obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance-concentration decay formula corresponding to the gas type, the heavy gas characteristic length, the deflagration warning point concentration, and the deflagration critical point concentration.

[0073] As an implementation approach, for gases heavier than air, such as hydrogen sulfide, the convection-diffusion equation in formula (5) above is no longer applicable. A heavy gas box model is used as the gas diffusion model to describe its concentration decay law:

[0074] (8)

[0075] In equation (8), This is the concentration at the lower explosive limit (equivalent to the deflagration critical point concentration, which can be taken as 4%, 40,000 ppm. Since the sensor range is 0-100 ppm, this is the point set by the algorithm). The distance from the concentration at the lower explosive limit (equivalent to the deflagration critical point concentration) Gas concentration at a given location (unit: ppm). The characteristic length of heavy gas, in meters, is determined by the tunnel width. (Most subway tunnel sections are 5m wide) and wind speed (Unit: m / s; hydrogen sulfide is both a flammable gas and highly toxic, therefore a conservative calculation is adopted, and wind direction is not considered in the calculation.) This is jointly determined by:

[0076] (9)

[0077] in, This is the wind speed coefficient, measured in seconds per meter (s / m); its value is 0.2.

[0078] Based on this, the concentration of the leaked gas at the lower explosive limit can be calculated. Decrease to target concentration (The distance at 100ppm, 0.01%, which is the upper limit of the sensor's measurement range) (Unit: meters), that is, the formula for the decrease in distance as a function of concentration for heavy gases is as follows:

[0079] (10)

[0080] Therefore, for gases heavier than air, such as hydrogen sulfide, this invention can calculate the longitudinal distance from the concentration at the lower explosion limit (deflagration threshold concentration) to the concentration at the deflagration warning point by replacing the target concentration in formula (10).

[0081] It should be noted that this invention sets the gas concentration corresponding to the deflagration warning zone of hydrogen sulfide to 100ppm-40000ppm (this range was selected by comprehensively considering the lower explosive limit of hydrogen sulfide, toxicity, and hardware sensors; the lower explosive limit of hydrogen sulfide is 40000ppm, and a dose of 100ppm can cause immediate death; the hydrogen sulfide gas detection sensor has a range of 0-100ppm). For hydrogen sulfide, this invention finds the concentration that satisfies 100ppm ≤ The longitudinal mileage range with a concentration of ≤ 40,000 ppm is designated as a deflagration warning zone. There is an extremely high risk of combustion, explosion, or toxicity at any point within this range.

[0082] This can be understood as follows: because the concentration of hydrogen sulfide gas at the actual leak point far exceeds the sensor's range, the sensor cannot directly measure it. This invention determines a virtual reference point (the starting point of the deflagration warning zone, 100 ppm) by extrapolating the concentration gradient. The hydrogen sulfide concentration at this point is defined as 100 ppm, and this concentration of hydrogen sulfide is sufficient to cause immediate death. Using this point as the starting point, the downstream concentration distribution is predicted using a distance-as-concentration decay formula, and based on this, the area with a concentration between 100 ppm and 40,000 ppm is designated as the deflagration warning zone (considering both toxicity and deflagration risk).

[0083] In some embodiments, the method for obtaining disaster identification results based on the longitudinal mileage of the explosion warning zone and the lateral offset distance among multiple target queue elements includes: obtaining the absolute lateral coordinates of the leak source based on the lateral offset distance of the nearest queue element among multiple target queue elements and the track centerline coordinates of the track where the track rescue vehicle is located; and combining the absolute lateral coordinates of the leak source and the longitudinal mileage of the explosion warning zone to obtain the disaster identification results.

[0084] Therefore, by determining the lateral offset distance using thermal imaging data, the absolute lateral coordinates of the leak source can be obtained, solving the problem that the railcar cannot directly measure the lateral gradient in the existing technology.

[0085] In some embodiments, the method further includes: obtaining high-temperature point identification results based on the temperature of each point in the thermal imaging data and the temperature safety threshold corresponding to each device in the tunnel; mapping the high-temperature point identification results to the longitudinal mileage of the tunnel to obtain ignition hazard zone identification results; and obtaining extremely high-risk zone identification results based on the ignition hazard zone identification results and the explosion warning zone. Ignition hazard zone identification here can be understood as detecting the surface temperature of electrical equipment; if the surface temperature is greater than or equal to the temperature safety threshold of the equipment, the longitudinal mileage of the ignition hazard zone is obtained.

[0086] As an example, the temperature safety thresholds for equipment are defined as follows: cable joints and junction boxes with surface temperatures >40℃, motors and lighting fixtures with surface temperatures >80℃, and any open flame or area with a temperature >200℃ (potentially an electrical short circuit or combustion). All high-temperature points exceeding the corresponding equipment safety thresholds are extracted from infrared thermal imaging images. These high-temperature points are then spatially mapped to tunnel mileage coordinates to obtain the ignition hazard zones. Extremely high risk area Defined as: This refers to the intersection of the explosion warning zone and the ignition hazard zone. If the intersection is not empty, the system outputs "There is an extremely high risk of explosion XX meters ahead, please stop the vehicle" and marks the area as a red flashing block on the electronic map. If the two are very close in space (less than 2 meters apart) but do not completely overlap, they are also treated as "extremely high risk" because the leak may be displaced due to airflow.

[0087] Therefore, this solution considers a joint early warning mechanism that takes into account both gas concentration conditions and ignition source conditions (whether there are electrical devices with abnormal temperatures) to improve the accuracy of early warnings and avoid missing potential high-risk scenarios.

[0088] Step S105: Based on the disaster identification results, issue an early warning.

[0089] In some embodiments, the method for obtaining early warning output information based on disaster identification results includes: outputting a warning stop instruction via voice when the ignition hazard zone identification result indicates the existence of an extremely high-risk zone; and forcibly stopping the vehicle when the distance between the rail rescue vehicle and the starting point of the explosion warning zone is less than a first distance threshold.

[0090] As an example, the following content is simultaneously displayed on the in-vehicle touchscreen and the rear command center screen: a longitudinal profile of the tunnel, indicating the current vehicle position; the location of the deflagration threshold (marked with a red cross), the predicted concentration decay curve (overlaid on the profile), the deflagration warning zone (a red semi-transparent strip), abnormal heat sources (i.e., high-temperature points, small red dots), and an extremely high-risk zone (bright red flashing). When the vehicle approaches the deflagration warning zone, the system automatically triggers a voice alarm: "Deflagration danger ahead, please stop," and slows down. When approaching the deflagration warning zone (≤2m), the vehicle is forced to stop. Special note: For hydrogen sulfide, as it is a highly toxic gas, as long as the sensor detects hydrogen sulfide at 7ppm (based on occupational safety standards), a bright red warning will be issued, and a voice alarm will be triggered: "Please wear respiratory protection equipment"; when approaching the hydrogen sulfide deflagration warning zone (≤2m), the vehicle is forced to stop.

[0091] The disaster identification method for rail rescue vehicles proposed in this application, by solving for the concentration gradient, can quickly determine the explosion warning zone using longitudinal mileage and gas concentration data sampled during the movement of the rescue vehicle along the track, even without pre-mapped conditions or a fixed sensor network. The method is simple, highly applicable, and solves the problem that existing solutions are not suitable for unfamiliar tunnels or scenarios without prior maps. Utilizing the inherent physical phenomena of high-pressure gas leaks, the lateral offset distance of the leak source relative to the track centerline can be directly obtained through thermal imaging data. Combined with the explosion warning zone, the location of the leak point can be identified, thereby determining high-risk areas and safety boundaries. This solution uses appropriate diffusion models to predict the concentration distribution of gases with different physical properties, such as methane and hydrogen sulfide, and delineates explosion warning zones accordingly, adapting to the hazardous characteristics of different leaking media. This solution implements a joint warning mechanism that simultaneously considers gas concentration conditions and ignition source conditions, improving warning accuracy while avoiding missed reports of potentially high-risk scenarios.

[0092] To clearly illustrate the above embodiments, specific examples will now be used for explanation.

[0093] Example 1: A track-based rescue vehicle encounters a methane leak inside a tunnel. The vehicle travels along the track at a speed of 5 km / h towards the leak point. After entering the affected area, the system detects that the methane concentration gradually increases from 0. The system records real-time sensing data such as longitudinal mileage and gas concentration every second, storing it in a 60-second sliding window. When the longitudinal concentration gradient of three consecutive points within the sliding window is positive, the system determines that the leak source is ahead. At this point, the system retrieves the most recent complete data interval (e.g., the longitudinal mileage and gas concentration of the five queue elements in this data interval are respectively...). =1200m =3%LEL; =1202m、 =4.5%LEL; =1204m =8%LEL; =1206m =10.5% LEL; =1208m =12.5%LEL), using least squares linear fitting, the average concentration gradient within this interval is approximately 1.25%LEL / m; taking the longitudinal mileage and gas concentration of the nearest point ( =1208m =12.5%LEL), calculated according to formula (3), we can obtain = 1208m + (30 - 12.5) / 1.25 = 1222m, meaning the starting point of the explosion warning zone (30% LEL) is located at longitudinal mileage 1222m. In the infrared thermal imaging image, a low-temperature zone with a temperature about 3°C ​​lower than the ambient temperature appears 0.2 meters to the right of the center of the image (corresponding to the right side of the track), thus determining that the leak source is located 0.2 meters to the right of the track. Assuming the wind speed measured by the micro weather station is 0.7 m / s, the wind direction is upwind of the rescue vehicle (downwind of the leak source), and the longitudinal diffusion coefficient is taken as 10 m² / s, the downstream concentration decay curve is calculated accordingly. When the concentration decays from 100% LEL to 30% LEL, the corresponding mileage distance is approximately 17m, meaning the explosion warning zone is from longitudinal mileage 1222m to longitudinal mileage 1239m. If the vehicle is traveling at mileage 1208m (starting from 1200m), the vehicle's thermal imaging suddenly shows that a cable connector 20m ahead (mileage 1228m) has a surface temperature of 48℃, significantly higher than the normal value (40℃). This location falls within the explosion warning zone (mileage 1222m to 1239m) and is therefore marked as an extremely high-risk area. The vehicle's screen will display a bright red flashing light and announce: "Extremely high risk of explosion 20 meters ahead, please stop." The command center at the rear will also receive the same information, allowing rescue personnel to stop in advance and prevent the vehicle from entering the explosion zone.

[0094] Example 2: Suppose a hydrogen sulfide leak occurs in a tunnel (density greater than air). The system detects an increase in hydrogen sulfide concentration and automatically switches to a heavy gas diffusion model, using formulas (9) and (10). The tunnel width is taken as 5 meters and the wind speed as 0.7 m / s. , ( (Approximately equal to 6), therefore, the longitudinal distance from the concentration of hydrogen sulfide at the lower explosive limit to the concentration at the explosion warning point is 68.2m, that is, the length of the hydrogen sulfide explosion warning zone is 68.2m. Combined with the concentration gradient method (5 points of continuous sampling), the coordinates of the 100ppm concentration point are calculated. Based on this, the system calculates the longitudinal coordinate range of the explosion warning zone (100ppm-40000ppm) and provides corresponding reminders to rescue personnel.

[0095] To achieve the above embodiments, this application also proposes a disaster identification device for a rail rescue vehicle. Figure 3 This is a schematic diagram of the structure of a disaster identification device for a rail rescue vehicle provided in an embodiment of this application. Figure 3 As shown, the disaster identification device for the rail rescue vehicle may include:

[0096] Among them, the data acquisition module 301 is used to acquire real-time sensing data collected by the sensing devices deployed on the rail rescue vehicle. The real-time sensing data includes longitudinal mileage, gas type, gas concentration, meteorological data and thermal imaging data.

[0097] The data processing module 302 is used to obtain the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image based on the thermal imaging data, and store the longitudinal distance, gas type, gas concentration, meteorological data and lateral offset distance as a queue element into the data queue.

[0098] The gradient processing module 303 is used to obtain the longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and longitudinal mileage in the data queue, and to obtain the most recently acquired target queue elements in the data queue when multiple consecutive longitudinal concentration gradients are positive.

[0099] The disaster identification module 304 is used to obtain the longitudinal mileage of the explosion warning zone based on the longitudinal mileage, gas type, gas concentration and meteorological data in multiple target queue elements, and to obtain the disaster identification result based on the longitudinal mileage of the explosion warning zone and the lateral offset distance in multiple target queue elements.

[0100] The early warning output module 305 is used to output early warnings based on the disaster identification results.

[0101] Furthermore, in one possible implementation of this application embodiment, when the gradient processing module 303 obtains the longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and longitudinal mileage in the data queue, it is used to:

[0102] Based on the gas concentration in the data queue, obtain the concentration difference between adjacent queue elements;

[0103] Based on the vertical mileage in the data queue, obtain the mileage difference between each adjacent queue element;

[0104] Based on the concentration difference and mileage difference of each adjacent queue element, the longitudinal concentration gradient of adjacent queue elements in the data queue is obtained.

[0105] Furthermore, in one possible implementation of this application embodiment, when the disaster identification module 304 obtains the longitudinal mileage of the explosion warning zone based on the longitudinal mileage, gas type, gas concentration, and meteorological data in multiple target queue elements, it is used to:

[0106] Obtain the gas type corresponding to the combustion and explosion warning point concentration and the deflagration critical point concentration from multiple target queue elements;

[0107] The average concentration gradient is obtained based on the longitudinal mileage and gas concentration in multiple target queue elements.

[0108] Based on the concentration at the explosion warning point, the average concentration gradient, and the longitudinal mileage and gas concentration of the nearest queue element among multiple target queue elements, the longitudinal mileage of the starting point of the explosion warning zone is obtained.

[0109] Based on the formula for distance decay with concentration corresponding to gas type, wind speed data in meteorological data, concentration at the explosion warning point and concentration at the deflagration critical point, the longitudinal distance from the leakage gas concentration to the explosion warning point concentration is obtained.

[0110] Based on the longitudinal distance and the longitudinal mileage of the starting point of the explosion warning zone, the longitudinal mileage of the ending point of the explosion warning zone is obtained.

[0111] Furthermore, in one possible implementation of this application embodiment, when the disaster identification module 304 obtains the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance decay formula corresponding to the gas type, wind speed data in meteorological data, the concentration at the deflagration warning point, and the concentration at the deflagration critical point, it is used to:

[0112] When the gas type is light gas, the longitudinal diffusion coefficient of the distance-concentration decay formula corresponding to the gas type is determined based on the wind direction data in the meteorological data.

[0113] Based on the formula for distance decay with concentration corresponding to gas type, longitudinal diffusion coefficient, wind speed data in meteorological data, concentration at the explosion warning point and concentration at the deflagration critical point, the longitudinal distance from which the concentration of leaked gas decreases from the deflagration critical point concentration to the explosion warning point concentration is obtained.

[0114] Furthermore, in one possible implementation of this application embodiment, when the disaster identification module 304 obtains the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance decay formula corresponding to the gas type, wind speed data in meteorological data, the concentration at the deflagration warning point, and the concentration at the deflagration critical point, it is used to:

[0115] When the gas type is heavy gas, the characteristic length of heavy gas is obtained based on the tunnel width of the tunnel where the rail rescue vehicle is located and the wind speed data in the meteorological data.

[0116] Based on the distance decay formula corresponding to the gas type with concentration, the characteristic length of heavy gas, the concentration at the combustion and explosion warning point and the concentration at the deflagration critical point, the longitudinal distance from which the concentration of leaked gas decreases from the concentration at the deflagration critical point to the concentration at the combustion and explosion warning point is obtained.

[0117] Furthermore, in one possible implementation of this application embodiment, the device further includes a risk area identification module, used for:

[0118] Based on the temperature of each point in the thermal imaging data and the corresponding temperature safety threshold of each device in the tunnel, the high temperature point identification result is obtained.

[0119] The high-temperature point identification results are mapped to the longitudinal mileage of the tunnel to obtain the ignition hazard zone identification results;

[0120] Based on the identification results of the ignition hazard zone and the explosion warning zone, the identification results of the extremely high risk zone were obtained.

[0121] Furthermore, in one possible implementation of this application embodiment, when the data processing module 302 obtains the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image based on the thermal imaging data, it is used to:

[0122] Based on thermal imaging data, the lateral pixel offset of the lowest temperature point in the thermal imaging image relative to the center line of the image is obtained.

[0123] The horizontal offset distance is obtained based on the horizontal pixel offset and pixel distance conversion coefficient.

[0124] Furthermore, in one possible implementation of this application embodiment, the warning output module 305 is used for:

[0125] If the ignition hazard zone identification result indicates the existence of an extremely high-risk area, a warning stop instruction will be output via voice.

[0126] When the distance between the rail rescue vehicle and the starting point of the explosion warning zone is less than the first distance threshold, the vehicle will be forced to stop.

[0127] Furthermore, in one possible implementation of this application embodiment, when the disaster identification module 304 obtains the disaster identification result based on the longitudinal mileage of the explosion warning zone and the lateral offset distance among multiple target queue elements, it is used for:

[0128] The absolute lateral coordinates of the leak source are obtained based on the lateral offset distance of the nearest queue element among multiple target queue elements and the centerline coordinates of the track where the track rescue vehicle is located.

[0129] By combining the absolute lateral coordinates of the leak source and the longitudinal mileage of the explosion warning zone, the disaster identification results are obtained.

[0130] It should be noted that the foregoing explanation of the disaster identification method for rail rescue vehicles also applies to the disaster identification device for rail rescue vehicles in this embodiment, and will not be repeated here.

[0131] In the foregoing descriptions of the embodiments, the terms "some embodiments," "examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0132] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0133] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for disaster identification using a rail rescue vehicle, characterized in that, include: The real-time sensing data collected by the sensing devices deployed on the track rescue vehicle includes longitudinal mileage, gas type, gas concentration, meteorological data, and thermal imaging data. Based on the thermal imaging data, the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image is obtained, and the longitudinal distance, gas type, gas concentration, meteorological data and the lateral offset distance are stored as a queue element in the data queue. Based on the gas concentration and longitudinal mileage in the data queue, the longitudinal concentration gradient of adjacent queue elements in the data queue is obtained, and if multiple consecutive longitudinal concentration gradients are positive, the most recently obtained target queue elements in the data queue are obtained. Based on the longitudinal mileage, gas type, gas concentration and meteorological data in the multiple target queue elements, the longitudinal mileage of the explosion warning zone is obtained, and based on the longitudinal mileage of the explosion warning zone and the lateral offset distance in the multiple target queue elements, the disaster identification result is obtained. Based on the disaster identification results, an early warning is issued.

2. The method for disaster identification of a rail rescue vehicle according to claim 1, characterized in that, The step of obtaining the longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and longitudinal mileage in the data queue includes: Based on the gas concentration in the data queue, obtain the concentration difference between adjacent queue elements; Based on the vertical mileage in the data queue, obtain the mileage difference between each adjacent queue element; Based on the concentration difference and mileage difference of each adjacent queue element, the longitudinal concentration gradient of adjacent queue elements in the data queue is obtained.

3. The method for disaster identification of a rail rescue vehicle according to claim 1, characterized in that, The process of obtaining the longitudinal mileage of the explosion warning zone based on the longitudinal mileage, gas type, gas concentration, and meteorological data from the multiple target queue elements includes: Obtain the combustion and explosion warning point concentration and the deflagration critical point concentration corresponding to the gas type in the multiple target queue elements; The average concentration gradient is obtained based on the longitudinal mileage and gas concentration in the multiple target queue elements; Based on the concentration at the explosion warning point, the average concentration gradient, and the longitudinal mileage and gas concentration of the nearest queue element among the multiple target queue elements, the longitudinal mileage of the starting point of the explosion warning zone is obtained. Based on the distance decay formula corresponding to the gas type, the wind speed data in the meteorological data, the concentration at the explosion warning point and the concentration at the deflagration critical point, the longitudinal distance from which the concentration of the leaked gas decreases from the concentration at the deflagration critical point to the concentration at the explosion warning point is obtained. Based on the longitudinal distance and the longitudinal mileage of the starting point of the explosion warning zone, the longitudinal mileage of the ending point of the explosion warning zone is obtained.

4. The method for disaster identification of a rail rescue vehicle according to claim 3, characterized in that, The method of obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance decay formula corresponding to the gas type, the wind speed data in the meteorological data, the concentration at the deflagration warning point, and the concentration at the deflagration critical point includes: When the gas type is light gas, the longitudinal diffusion coefficient of the distance-as-concentration decay formula corresponding to the gas type is determined based on the wind direction data in the meteorological data. Based on the distance-concentration decay formula corresponding to the gas type, the longitudinal diffusion coefficient, the wind speed data in the meteorological data, the concentration at the explosion warning point, and the concentration at the deflagration critical point, the longitudinal distance from which the leaked gas concentration decreases from the deflagration critical point concentration to the explosion warning point concentration is obtained.

5. The method for disaster identification of a rail rescue vehicle according to claim 3, characterized in that, The method of obtaining the longitudinal distance from the deflagration critical point concentration to the deflagration warning point concentration based on the distance decay formula corresponding to the gas type, the wind speed data in the meteorological data, the concentration at the deflagration warning point, and the concentration at the deflagration critical point includes: When the gas type is heavy gas, the heavy gas characteristic length is obtained based on the tunnel width of the tunnel where the rail rescue vehicle is located and the wind speed data in the meteorological data. Based on the distance decay formula corresponding to the gas type with concentration, the characteristic length of the heavy gas, the concentration of the combustion and explosion warning point, and the concentration of the deflagration critical point, the longitudinal distance from which the concentration of the leaked gas decreases from the concentration of the deflagration critical point to the concentration of the combustion and explosion warning point is obtained.

6. The method for disaster identification of a rail rescue vehicle according to claim 1, characterized in that, The method further includes: Based on the temperature of each point in the thermal imaging data and the temperature safety threshold corresponding to each device in the tunnel, the high temperature point identification result is obtained. The high-temperature point identification results are mapped to the longitudinal mileage of the tunnel to obtain the ignition hazard zone identification results; Based on the ignition hazard zone identification results and the explosion warning zone, the extremely high risk zone identification results are obtained.

7. The method for disaster identification of a rail rescue vehicle according to claim 1, characterized in that, The step of obtaining the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image based on the thermal imaging data includes: Based on the thermal imaging data, the lateral pixel offset of the lowest temperature point in the thermal imaging image relative to the center line of the image is obtained. The horizontal offset distance is obtained based on the horizontal pixel offset and pixel distance conversion coefficient.

8. The method for disaster identification of a rail rescue vehicle according to claim 6, characterized in that, Based on the disaster identification results, early warning output information is obtained, including: If the ignition hazard zone identification result indicates the existence of an extremely high-risk zone, a warning stop instruction will be output via voice. When the distance between the rail rescue vehicle and the starting point of the explosion warning zone is less than a first distance threshold, the vehicle will be forced to stop.

9. The method for disaster identification of a rail rescue vehicle according to claim 1, characterized in that, The disaster identification result is obtained based on the longitudinal mileage of the explosion warning zone and the lateral offset distance among the multiple target queue elements, including: The absolute lateral coordinates of the leak source are obtained based on the lateral offset distance of the nearest queue element among the multiple target queue elements and the centerline coordinates of the track where the track rescue vehicle is located. By combining the absolute lateral coordinates of the leak source and the longitudinal mileage of the explosion warning zone, the disaster identification result is obtained.

10. A disaster identification device for a rail rescue vehicle, characterized in that, include: The data acquisition module is used to acquire real-time sensing data collected by the sensing devices deployed on the rail rescue vehicle. The real-time sensing data includes longitudinal mileage, gas type, gas concentration, meteorological data, and thermal imaging data. The data processing module is used to obtain the lateral offset distance of the lowest temperature point in the thermal imaging image relative to the center line of the image based on the thermal imaging data, and store the longitudinal distance, gas type, gas concentration, meteorological data and the lateral offset distance as a queue element into the data queue. The gradient processing module is used to obtain the longitudinal concentration gradient of adjacent queue elements in the data queue based on the gas concentration and longitudinal mileage in the data queue, and to obtain the most recently acquired target queue elements in the data queue when multiple consecutive longitudinal concentration gradients are positive. The disaster identification module is used to obtain the longitudinal mileage of the explosion warning zone based on the longitudinal mileage, gas type, gas concentration and meteorological data in the multiple target queue elements, and to obtain the disaster identification result based on the longitudinal mileage of the explosion warning zone and the lateral offset distance in the multiple target queue elements. The early warning output module is used to output early warnings based on the disaster identification results.

Images