Method and system for comprehensive pipeline corridor gas pipeline leakage explosion accident risk classification and early warning

Abstract

This application relates to the field of urban underground engineering technology, and provides a method and system for risk classification and early warning of gas pipeline leakage and explosion accidents in integrated utility tunnels. It simulates gas pipeline leakage and explosion accidents in integrated utility tunnels, divides the tunnels into different accident zones based on the simulation results, and simultaneously collects spatial parameters and temporal pressure data from various measuring points during the simulation process to obtain feature vectors identifying the included areas of different accident zones. Furthermore, it fuses these feature vectors to obtain the peak overpressure and pressure rise time at each measuring point during a gas pipeline leakage and explosion in the integrated utility tunnel. This enables a comprehensive risk zoning assessment of gas pipeline leakage and explosion accidents in integrated utility tunnels, forming a dynamic zoning assessment model for gas explosion risk in integrated utility tunnels. It classifies and categorizes dynamic explosion overpressure damage, effectively solving the problems of rapid extraction of gas leakage and explosion characteristics and efficient assessment of explosion risks in confined underground spaces.

Description

Technical Field

[0001] This application relates to the field of urban underground engineering technology, and in particular to a risk classification and early warning method and system for gas pipeline leakage and explosion accidents in integrated utility tunnels. Background Technology

[0002] In urban underground utility tunnels and other infrastructure, the risk of pipeline leaks and explosions is extremely high, and traditional gas monitoring and early warning systems face significant technical bottlenecks.

[0003] Gas pipelines provide the gas source for gas explosions, and the flammable nature of high-voltage power lines provides the ignition source. However, long, narrow underground tunnels can easily accumulate large amounts of gas, which can then rapidly spread, expanding the affected area. Furthermore, the presence of numerous high-risk urban lifelines (gas, electricity, HVAC, water supply, etc.) within these confined underground spaces further complicates gas explosions. For example, power lines can easily cause electrical fires and high-voltage explosions, while HVAC lines can result in high-temperature surfaces. Additionally, open flame work by personnel and accidental static electricity within these underground spaces can all become sources of ignition. In these long, confined underground spaces stretching for tens of kilometers, the explosive energy generated after an explosion is difficult to release, leading to a massive energy accumulation and extremely dangerous consequences. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for risk classification and early warning of gas pipeline leakage accidents in integrated utility tunnels, so as to solve or alleviate the problems existing in the above-mentioned prior art.

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] This application provides a risk classification and early warning method for gas pipeline leakage and explosion accidents in integrated utility tunnels, including:

[0007] Accident simulations of gas pipeline leaks and explosions in integrated utility tunnels were conducted using an accident simulation device. Based on the simulation results, the integrated utility tunnels were divided into different accident zones. Simultaneously, data was collected during the accident simulation process. Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing area identifiers for different accident zones are obtained; among them, It is a positive integer;

[0008] By fusing the feature vectors of the included area identifiers from different accident zones, the results of gas pipeline leaks and explosions in integrated utility tunnels can be obtained. The overpressure peak value and pressure rise time at each measuring point.

[0009] Preferably, the leakage volume rate is obtained through accident simulation. and duration of leakage According to the gas leak diffusion model:

[0010]

[0011] During a gas pipeline leak and explosion, the integrated utility tunnel is divided into a gas-bearing zone and a non-gas-bearing zone; where, This is the distance from the gas leak point to the boundary of the area containing the gas. The gas diffusion coefficient during a gas pipeline leak and explosion. This is a correction factor for the gas diffusion range under the ventilation conditions of a utility tunnel;

[0012] The non-gas-filled zone is divided into an explosion reflection zone and an explosion open zone by using the spatial parameters of the weak surface layout.

[0013] Preferably, a fusion model based on the characteristics of gas-filled areas is adopted:

[0014]

[0015] Feature fusion is performed on the feature vectors of gas-filled zones that include area identifiers.

[0016] In the formula, For input characteristics of areas with gas supply, This is the feature fusion weight matrix for areas with gas supply; This is a spatial feature vector representing the distance from the measuring point in the gas-bearing area to the gas boundary, including the straight-line distance from the measuring point to the gas boundary. ; Features of the explosion source, including the energy characteristics of the explosion source. Explosion source area characteristics ; Characteristics of gas, For regional identification features, For the time-series pressure at the measuring point,

[0017] This is the feature encoding bias term for the gas-zone feature fusion model. The spatial attenuation factor for areas with gas combustion;

[0018] It has the characteristics of a gas-fired zone; The energy density of the explosion source.

[0019] Preferably, the overpressure peak prediction model for gas-filled areas is used:

[0020]

[0021] Predicting the peak overpressure in areas with gas combustion;

[0022] In the formula, For the predicted overpressure peak in the gas-bearing area, The output layer weights of the overpressure peak prediction model for gas-filled areas are... Let be the activation function of the overpressure peak prediction model for gas-filled areas. The hidden layer weights are for the overpressure peak prediction model in the gas-filled zone. It has the characteristics of a gas-fired zone. This is the hidden layer bias term of the overpressure peak prediction model for gas-filled areas. This is the output layer bias term of the overpressure peak prediction model for gas-filled areas, where C represents the gas concentration. This is a correction term for gas concentration. This is a correction term for the energy source of the explosion. It is the energy source of the explosion.

[0023] Preferably, a fusion model based on the characteristics of the open blast zone is used:

[0024]

[0025] Feature fusion is performed on the feature vectors of the open blast zone that include area identifiers;

[0026] In the formula, The characteristics of the fusion of the open zone of the explosion; The feature fusion weight matrix for the open zone of the explosion. This is the spatial feature vector of the distance from the measuring point in the explosion open zone to the gas boundary; This refers to the overpressure characteristics at the gas boundary of the open explosion zone. Characteristics of gas, For regional identification features, The straight-line distance from the measuring point in the open explosion zone to the gas boundary; This is the distance attenuation factor for the open zone of the explosion.

[0027] Preferably, the overpressure peak prediction model for the open explosion zone is used:

[0028]

[0029] Predicting the peak overpressure in the open zone of an explosion;

[0030] In the formula, The predicted peak overpressure in the open zone of the explosion. The output layer weights are the overpressure peak prediction model for the open explosion zone. is the activation function for the overpressure peak prediction model in the open explosion zone. The hidden layer weights are used in the overpressure peak prediction model for the open explosion zone. The characteristics of the fusion of the open zone of the explosion, This represents the hidden layer bias term in the overpressure peak prediction model for the open explosion zone. This is the output layer bias term of the overpressure peak prediction model for the open explosion zone.

[0031] Preferably, a fusion model based on the characteristics of the explosion reflection zone is used:

[0032]

[0033] Feature fusion is performed on the feature vectors containing area identifiers in the explosion reflection zone;

[0034] In the formula, To employ a cyclic network structure to study the temporal characteristics of the explosion reflection zone Bimodal dynamic features obtained by sequence feature extraction; The time interval between the two pressure peaks in the time-series pressure sequence of the explosion reflection zone;

[0035] The fusion characteristics of the explosion reflection zone The characteristic combination weights of the explosion reflection zone, This is the spatial feature vector of the distance from the measuring point in the explosion reflection zone to the gas boundary. This refers to the overpressure characteristics at the gas boundary of the explosion reflection zone. Characteristics of gas, For regional identification features, The distance is the straight-line distance from the measuring point in the explosion reflection zone to the gas boundary. The distance is the straight-line distance from the measuring point in the explosion reflection zone to the normally closed fire door inside the integrated utility tunnel. This represents the straight-line distance from the measuring point in the explosion reflection zone to the turning point inside the integrated utility tunnel. This is the straight-line distance from the measuring point in the explosion reflection zone to the branching point inside the integrated utility tunnel.

[0036] Preferably, according to the overpressure peak prediction model of the explosion reflection zone:

[0037]

[0038] Predicting the peak overpressure in the explosion reflection zone;

[0039] In the formula, These are the predicted peak values ​​of incident pressure and reflected pressure in the explosion reflection zone, respectively. The output layer weights are the overpressure peak prediction model for the explosion reflection zone. is the activation function for the overpressure peak prediction model in the explosion reflection zone. The hidden layer weights are used in the overpressure peak prediction model for the explosion reflection zone. The fusion characteristics of the explosion reflection zone This represents the hidden layer bias term in the overpressure peak prediction model for the explosion reflection zone. This is the output layer bias term of the overpressure peak prediction model for the explosion reflection zone;

[0040] The distance is the straight-line distance from the measuring point in the explosion reflection zone to the normally closed fire door inside the integrated utility tunnel. This represents the straight-line distance from the measuring point in the explosion reflection zone to the turning point inside the integrated utility tunnel. The straight-line distance from the measuring point in the explosion reflection zone to the bifurcation point inside the integrated utility tunnel;

[0041] The predicted overpressure peak value for the explosion reflection zone.

[0042] Preferably, according to the boost time prediction model:

[0043]

[0044] Predict the boost time for different accident zones;

[0045] In the formula, For the predicted pressure rise time in areas with gas supply, These are the output layer weights and bias terms of the gas-filled zone pressure rise time prediction model, respectively. The activation function for the gas-fired zone pressure rise time prediction model is... These are the hidden layer weights and bias terms of the gas-filled zone pressure rise time prediction model; This represents the fusion characteristics of the gas-bearing zone; C represents the gas concentration. This is a correction term for gas concentration. This is a correction term for the energy of the explosion source; is the activation function for the gas-fired zone pressurization time prediction model;

[0046] The predicted pressurization time for the open blast zone. These are the output layer weights and bias terms of the explosion open zone pressurization time prediction model; These are the hidden layer weights and bias terms of the explosion open zone pressurization time prediction model, respectively; The characteristics of the fusion of the open zone of the explosion; The activation function for the explosion open zone pressurization time prediction model;

[0047] Predicted pressure rise time for measuring points in the explosion reflection zone; Features of the fusion of the explosion reflection zone; , These are the hidden layer weights and bias terms of the explosion reflection zone pressurization time prediction model, respectively. , These are the output layer weights and bias terms of the explosion reflection zone pressurization time prediction model; The fusion characteristics of the explosion reflection zone; is the activation function for the explosion reflection zone pressurization time prediction model.

[0048] This application embodiment also provides a risk classification and early warning system for gas pipeline leaks and explosions in integrated utility tunnels, which uses any of the above-described risk classification and early warning methods for gas pipeline leaks and explosions in integrated utility tunnels to predict gas pipeline leaks and explosions in integrated utility tunnels. The system includes:

[0049] The simulation and data acquisition unit is configured to simulate gas pipeline leaks and explosions in the integrated utility tunnel using an accident simulation device, and to divide the integrated utility tunnel into different accident zones based on the simulation results; simultaneously, it collects data during the accident simulation process. Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing area identifiers for different accident zones are obtained; among them, It is a positive integer;

[0050] The leak prediction unit is configured to perform multi-feature fusion on the feature vectors containing area identifiers from different accident zones to obtain the leak prediction result for gas pipeline leaks and explosions in integrated utility tunnels. The overpressure peak value and pressure rise time at each measuring point.

[0051] Beneficial effects:

[0052] The risk classification and early warning method and system for gas pipeline leakage and explosion accidents in integrated utility tunnels provided in this application embodiment uses an accident simulation device to simulate gas pipeline leakage and explosion accidents in integrated utility tunnels. Based on the simulation results, the integrated utility tunnel is divided into different accident zones, and data is collected during the accident simulation process. By analyzing the spatial parameters and temporal pressure at each measuring point, feature vectors identifying the included areas of different accident zones are obtained. Furthermore, these feature vectors are fused using multiple features to determine the characteristics of gas pipeline leaks and explosions within the integrated utility tunnel. By measuring the overpressure peak value and pressure rise time at each measuring point, a comprehensive risk zoning assessment of gas explosion accidents in integrated utility tunnels is achieved. This forms a dynamic zoning assessment model for gas explosion risks in integrated utility tunnels, classifying and grading dynamic explosion overpressure damage. This effectively solves the problems of rapid extraction of gas leak explosion characteristics and efficient assessment of explosion risks in narrow, confined underground spaces, providing support for the safe operation of urban underground integrated utility tunnels and the safety of gas applications. It also provides gas explosion disaster risk alerts and different emergency response measures for the operation of the integrated utility tunnel central control platform. Attached Figure Description

[0053] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. Wherein:

[0054] Figure 1 This is a flowchart illustrating a risk classification and early warning method for gas pipeline leakage and explosion accidents in an integrated utility tunnel, according to some embodiments of this application.

[0055] Figure 2 This is a logical schematic diagram of a risk classification and early warning method for gas pipeline leakage and explosion accidents in integrated utility tunnels provided according to some embodiments of this application;

[0056] Figure 3 This is a logical diagram illustrating a zoned early warning system for gas pipeline leaks and explosions in an integrated utility tunnel, provided according to some embodiments of this application.

[0057] Figure 4 This is a schematic diagram of a gas pipeline leakage and explosion zoning structure for an integrated utility tunnel according to some embodiments of this application;

[0058] Figure 5 This is a comparative diagram of explosion overpressure on both sides of an explosion source in a gas-filled zone, provided according to some embodiments of this application.

[0059] Figure 6 This is a schematic diagram of the explosion overpressure in an open explosion zone according to some embodiments of this application;

[0060] Figure 7 This is a schematic diagram of the explosion overpressure in the explosion reflector zone according to some embodiments of this application;

[0061] Figure 8 This is a logical schematic diagram of a risk classification and early warning method for gas pipeline leakage and explosion accidents in integrated utility tunnels provided according to some embodiments of this application;

[0062] Figure 9 This is a schematic diagram of the structure of a risk classification and early warning system for gas pipeline leakage and explosion accidents in an integrated utility tunnel, provided according to some embodiments of this application. Detailed Implementation

[0063] The present application will now be described in detail with reference to the accompanying drawings and embodiments. Various examples are provided by way of explanation and not by way of limitation. In fact, those skilled in the art will understand that modifications and variations can be made to the present application without departing from the scope or spirit of the present application. For example, a feature shown or described as part of one embodiment may be used in another embodiment to produce yet another embodiment. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention should fall within the scope of protection of the embodiments of the present invention.

[0064] In long, narrow underground spaces stretching tens of kilometers, gas leaks spread rapidly. Without accurately capturing the characteristics of both the gas leak's diffusion and the explosion source, the explosion location cannot be precisely determined, the explosion risk cannot be accurately predicted, and risk visualization becomes difficult. Furthermore, explosions are rare, high-severity, and difficult-to-respond-to-emergency events, making risk visualization even more challenging. Existing risk warning systems for underground space explosions primarily employ three-dimensional prediction models, which are inefficient and ill-suited for long, narrow spaces spanning tens of kilometers.

[0065] Based on this, this embodiment provides a risk classification and early warning method for gas pipeline leakage and explosion accidents in integrated utility tunnels. By comprehensively assessing the risk zoning of gas explosion accidents in integrated utility tunnels, a dynamic zoning assessment model for gas explosion risks in integrated utility tunnels is formed. Dynamic explosion overpressure damage is zoned and classified, effectively solving the problems of rapid extraction of gas leakage and explosion characteristics and efficient assessment of explosion risks in narrow, confined underground spaces. Figures 1 to 7 As shown, the method includes:

[0066] Step S101: Conduct an accident simulation of a gas pipeline leak and explosion in the integrated utility tunnel using an accident simulation device. Based on the simulation results, divide the integrated utility tunnel into different accident zones. Simultaneously, collect data during the accident simulation process. Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing the regional identifiers of different accident zones are obtained.

[0067] In this embodiment, an accident simulation device is used to simulate a gas pipeline leak and explosion in a utility tunnel. This device serves as a scaled-down accident data source for predicting overpressure in gas explosions in utility tunnels, providing training data for accident simulation assessment. Specifically, it includes: a utility tunnel shell structure, a gas leak module, an explosion source simulation module, and a signal detection module. These modules act as physical experimental models, generating labeled data for "leakage characteristics - explosion source characteristics - measuring point location - explosion characteristics," including at least the gas distribution location, explosion source location, measuring point location, explosion source energy, and area.

[0068] Among them, the multi-configuration pipe gallery shell structure is scaled down to a 1:20 scale based on the actual pipe gallery structure, with cross-sectional dimensions of [missing information]. The straight pipe gallery model is long A multi-configuration pipe gallery shell structure includes at least multiple geometric simulation shells with multiple routes, multiple pipelines, and multiple facilities. Different structural routes include straight pipe gallery models, turning models, bifurcation models, and slope models. The angle of the turning model includes... The bifurcation model includes T-bifurcation and cross bifurcation; the slope model is implemented using the turning model; the angles of the uphill and downhill slopes include... Different pipe gallery layouts, leakage conditions, and leakage times collectively influence the gas diffusion pattern, while different gas distribution ranges and morphologies affect the explosion propagation pattern. Furthermore, different pipe gallery structural layouts also affect the reflection path of explosion energy, thus influencing the explosion propagation pattern.

[0069] The gas leak detection module includes a gas pipeline that allows gas to pass through, a gas leak control solenoid valve, and a gas cylinder. The gas pipeline is installed at intervals of... A circular leak hole is provided in the gas pipeline to control the gas flow pressure. The explosion source simulation module is used in the utility tunnel model at intervals of... An electric spark igniter is installed, which can adjust the energy of the electrostatic spark, instantly igniting the gas upon activation. The signal detection module includes a gas sensor and a pressure sensor, both of which are positioned along the pipe gallery shell model at intervals of... One unit is installed, with the gas sensor located on the top of the pipe rack housing module and the pressure sensor located on the side wall of the pipe rack housing module.

[0070] Accident scenario reconstruction is performed using an accident simulation device to obtain explosion accident scenario data. The gas leak module and explosion source module are adjusted to simulate and select target gas leak characteristics and explosion source characteristics, thereby acquiring a large amount of leakage and explosion data under accident scenarios. In this embodiment, the spatial characteristics of gas explosion overpressure are divided into a gas zone (Zone A) and a non-gas zone (Zone B). The non-gas zone (Zone B) is further divided into an explosion open zone (Zone B1) and an explosion reflection zone (Zone B2) according to the pipe gallery structure.

[0071] Accident simulation devices are used to conduct accident scenario planning, including area delineation and multi-dimensional feature collection. Area delineation divides the gas leak diffusion characteristics into a gas-bearing zone (Zone A) and a non-gas-bearing zone (Zone B), while simultaneously collecting preset data. Spatial parameters (corridor structure, blast source location, blast source characteristics, measuring point location, etc.) of each measuring point and temporal pressure are used to generate a region identifier (Area A identifier). B1 area sign B2 area sign ) eigenvectors.

[0072] In a specific example, the leakage volume rate obtained through accident simulation. and duration of leakage According to the gas leak diffusion model:

[0073]

[0074] During a gas pipeline leak and explosion, the integrated utility tunnel is divided into a gas-bearing zone and a non-gas-bearing zone; where, This is the distance from the gas leak point to the boundary of the area containing the gas. The gas diffusion coefficient during a gas pipeline leak and explosion. This is a correction factor for the gas diffusion range under the ventilation conditions of a utility tunnel, and its value is [value missing]. Furthermore, the non-gas-filled area is divided into an explosion-reflecting zone and an explosion-open zone based on the spatial parameters of the weak-side layout (fire doors). Specifically, the normally closed direction of the fire doors constitutes the explosion-reflecting zone, while the normally open direction is the explosion-open zone.

[0075] For the gas-filled zone (Zone A), the explosion propagation is influenced by a combination of factors including gas distribution, explosion source characteristics, and the spatial location of the measuring point. Key characteristics include the relative position of the measuring point to the gas and the explosion source, the explosion source energy, and the explosion source area. For the open explosion zone (Zone B1), there is no gas between the gas boundary and the normally open fire door. Key characteristics include the distance from the gas zone boundary to the measuring point and the explosion overpressure characteristics at the gas boundary. For the explosion reflection zone (Zone B2), there is no gas between the gas boundary and the normally closed fire door, turning point, and bifurcation wall. Key characteristics include the relative position of the measuring point to the fire door, turning point, bifurcation point, and gas zone boundary, and the explosion overpressure characteristics at the gas boundary. Due to wall reflection, the pressure overpressure curve exhibits a distinct bimodal structure. The time interval between the two pressure peaks (bimodal structure characteristic) is affected by the relative distance between the measuring point and the fire door and the gas boundary.

[0076] Step S102: Perform multi-feature fusion on the feature vectors containing area identifiers of different accident zones to obtain the results of gas pipeline leaks and explosions in integrated utility tunnels. The overpressure peak value and pressure rise time at each measuring point. In this embodiment, feature vectors containing region identifiers from different zones are input into a neural network model for multi-feature fusion. For the gas-filled zone (Zone A), this area is filled with gas, and the explosion propagation is affected by the coupling of multiple factors such as "gas distribution, explosion source characteristics, and spatial location of measuring points". The focus is on capturing the synergistic effect of "gas combustion - explosion source energy - spatial attenuation" to effectively avoid prediction bias caused by a single feature. Specifically, according to the feature fusion model of the gas-filled zone (Zone A):

[0077]

[0078] Feature fusion is performed on the feature vectors of the gas-filled zone (Zone A) that include area identifiers;

[0079] In the formula, The input characteristics are for the gas-filled area (Area A). This is the feature fusion weight matrix for the gas-bearing area (Area A); This is a spatial feature vector representing the distance from the measuring point in the gas-bearing zone (Zone A) to the gas boundary, including the straight-line distance from the measuring point to the gas boundary. ; Features of the explosion source, including the energy characteristics of the explosion source. Explosion source area characteristics ; Characteristics of gas, For regional identification features, The time-series pressure is measured at the gas-bearing zone (Zone A).

[0080] This refers to the bias term in the feature encoding of the prediction model for the gas-bearing area (Area A). The spatial attenuation factor is the measurement point closer to the gas boundary in the gas-bearing zone (Zone A). The closer the value is to 1, the more significant the enhancement effect on the feature encoding bias term. The farther the measurement point is from the gas boundary, the greater the spatial attenuation factor. It exhibits exponential decay; It has the characteristics of a gas-fired zone (Zone A); The energy density of the explosion source (energy-area conversion factor).

[0081] Furthermore, according to the overpressure peak prediction model for gas-filled areas:

[0082]

[0083] Predict the peak overpressure in the gas-bearing zone (Zone A); where, The predicted overpressure peak value is for the gas-bearing area (Area A). The output layer weights of the overpressure peak prediction model for gas-filled areas are... The ReLU function is used as the activation function for the overpressure peak prediction model in the gas-filled zone. The hidden layer weights are for the overpressure peak prediction model in the gas-filled zone. It has the characteristics of a gas-fired zone. This is the hidden layer bias term of the overpressure peak prediction model for gas-filled areas. This is the output layer bias term of the overpressure peak prediction model for gas-filled areas, where C represents the gas concentration. This is a gas concentration correction term, which indicates that the higher the proportion of combustible components, the larger the value of the correction term, and the more significant the compensation effect on the overpressure peak. This is a correction term for the explosion source energy, indicating that the greater the total energy of the explosion source, the more significant the compensation effect on the overpressure peak. It is the energy source of the explosion.

[0084] The open explosion zone (Zone B1) is located between the gas boundary and the normally open fire door. There is no gas present. The explosion overpressure is solely due to the unidirectional propagation and natural attenuation of the explosion shock wave from the gas-containing zone (Zone A). Key characteristics include the distance from the boundary of the gas-containing zone (Zone A) to the measuring point and the explosion overpressure characteristics at the gas boundary. Specifically, according to the explosion open zone characteristic fusion model:

[0085]

[0086] Feature fusion is performed on the feature vectors containing area identifiers of the blast open zone (B1 area). Where, The fusion characteristics of the blast open zone (B1 area); The feature fusion weight matrix for the blast open zone (B1 zone) This is the spatial characteristic vector of the distance from the measuring point in the explosion open zone (B1 zone) to the gas boundary; The overpressure characteristics at the gas boundary of the open explosion zone (B1 zone) are as follows. Characteristics of gas, For regional identification features, The straight-line distance from the measuring point in the explosion open zone (B1 zone) to the gas boundary; The distance attenuation factor is the distance to the open blast zone (B1 zone).

[0087] The explosion reflection zone (Zone B2) is located between the gas boundary and the normally closed fire door, turning wall, and bifurcation wall. No gas is present there. The explosion overpressure is formed by the superposition of the incident wave and the reflected wave from the wall, exhibiting a distinct bimodal structure. Key characteristics include the relative positions of the measuring point to the fire door, turning point, bifurcation point, and gas zone boundary, as well as the explosion overpressure characteristics at the gas boundary. The explosion reflection zone (Zone B2) characteristic fusion model is as follows:

[0088]

[0089] Feature fusion is performed on the feature vector containing the region identifier of the explosion reflection zone (B2 zone); where, To employ a cyclic network structure to study the temporal characteristics of the explosion reflection zone (B2 zone) Bimodal dynamic features obtained by sequence feature extraction; The time interval between the two pressure peaks in the time-series pressure sequence of the explosion reflection zone (B2 zone).

[0090] The fusion characteristics of the explosion reflection zone (B2 zone) The characteristic combination weights of the explosion reflection zone (B2 zone) are... This is the spatial characteristic vector of the distance from the measuring point in the explosion reflection zone (B2 zone) to the gas boundary. The overpressure characteristics at the gas boundary of the explosion reflection zone (B2 zone) are... Characteristics of gas, For regional identification features, The distance is the straight-line distance from the measuring point in the explosion reflection zone (Zone B2) to the gas boundary. This refers to the straight-line distance from the measuring point in the explosion reflection zone (Zone B2) to the normally closed fire door inside the integrated utility tunnel. The distance is the straight-line distance from the measuring point in the explosion reflection zone (B2 zone) to the turning point inside the integrated utility tunnel. The distance is the straight-line distance from the measuring point in the explosion reflection zone (B2 zone) to the bifurcation point inside the integrated utility tunnel.

[0091] In this embodiment, through To correct the bimodal characteristic, the larger the interval between the two peaks, the weaker the superposition effect of the reflected and incident waves. The smaller the value, the stronger the cumulative effect. The larger. Through Effectively reflects the reflection superposition effect of the explosion reflection zone (B2 zone), and measures the coupling relationship between the "incident wave base intensity" and the "superposition intensity of multiple reflection sources". The larger the value, the higher the proportion of the emitted wave's contribution to the total overpressure. The smaller the value, the closer the measuring point is to the three types of reflection sources, the shorter the propagation path of the reflected wave, the less energy attenuation, and the stronger the superposition effect of multiple reflection sources.

[0092] Furthermore, according to the overpressure peak prediction model of the explosion reflection zone (B2 zone):

[0093]

[0094] Predict the peak overpressure in the explosion reflection zone (B2 zone);

[0095] In the formula, These are the predicted peak values ​​of incident pressure and reflected pressure in the explosion reflection zone (B2 zone), respectively. The output layer weights are the overpressure peak prediction model for the explosion reflection zone. The activation function of the overpressure peak prediction model for the explosion reflection zone is used. This effectively avoids Ford's behavior when the gradient value is small and effectively adapts to the complex dynamic changes of bimodal characteristics. Specifically, the LeakyReLU function is used. The hidden layer weights are used in the overpressure peak prediction model for the explosion reflection zone. The fusion characteristics of the explosion reflection zone This represents the hidden layer bias term in the overpressure peak prediction model for the explosion reflection zone. This is the output layer bias term of the overpressure peak prediction model for the explosion reflection zone.

[0096] This refers to the straight-line distance from the measuring point in the explosion reflection zone (Zone B2) to the normally closed fire door inside the integrated utility tunnel. The distance is the straight-line distance from the measuring point in the explosion reflection zone (B2 zone) to the turning point inside the integrated utility tunnel. The straight-line distance from the measuring point in the explosion reflection zone (B2 zone) to the bifurcation point inside the integrated utility tunnel; The predicted overpressure peak value for the explosion reflection zone (B2 zone). (This is achieved through...) The reflection source with the strongest influence on the overpressure at the measuring point is extracted, and its spatial constraint strength is quantified. The smaller the value, the higher the total overpressure peak. There are three types of reflection sources in the explosion reflection zone (B2 zone). The reflection source closest to the measuring point has the most significant impact on the overpressure.

[0097] In this embodiment, pressure-time series data is collected during accident simulation using the accident simulation device, and the time point corresponding to the initial static pressure of each measuring point is determined by baseline calibration of the pressure-time curve of each measuring point. The time and space at which the pressure begins to rise rapidly in the pressure-time curve are determined to be the initial arrival time of the incident wave at the measuring point in the accident simulation. (Among them, when the pressure change exceeds the normal range of data noise fluctuation, it is determined to be the initial arrival of the incident wave).

[0098] For areas with gas ( (area) and open blast zone ( (Region), the pressure-time curve usually exhibits a single-peak structure, with the maximum peak value being the incident peak value, and its corresponding time marker being... The pressure rise time at the measuring point for:

[0099]

[0100] That is, the time it takes for the pressure to rise from the initial incident wave to the peak value.

[0101] For the explosion reflection zone ( Due to the reflection effect of structural surfaces, the pressure-time series data exhibits a bimodal structure with incident and reflected peaks. Therefore, the pressure rise time is used to characterize the pressure rise process when the incident wave first arrives at the measuring point. The first peak value (first peak value) in the pressure-time curve is used as the criterion for determining the pressure rise time, unaffected by the reflected peak value. The time corresponding to the first peak value in the pressure-time curve is marked as... The pressurization time of the explosion reflection zone Defined as:

[0102]

[0103] Furthermore, the pressure rise time at each measuring point during the accident simulation process was analyzed. The data is labeled and combined with the fusion features of the corresponding measurement points to form the training dataset for the boost time prediction model. The constructed boost time prediction model based on the neural network architecture is trained, and the model parameters are updated by minimizing the mean square error loss function. The boost time prediction models for different accident zones use independent network parameters and activation functions.

[0104] Specifically, through the constructed boost time prediction model:

[0105]

[0106] Predict the boost time for different accident zones.

[0107] In the formula, in the formula, For the predicted pressure rise time in areas with gas supply, These are the output layer weights and bias terms of the gas-filled zone pressure rise time prediction model, respectively. The activation function for the gas-fired zone pressure rise time prediction model is... These are the hidden layer weights and bias terms of the gas-filled zone pressure rise time prediction model; This represents the fusion characteristics of the gas-bearing zone; C represents the gas concentration. This is a correction term for gas concentration. This is a correction term for the energy of the explosion source; The activation function for the gas-fired zone pressurization time prediction model is preferably the same as that used in the gas-fired zone overpressure peak prediction model. This is a mapping function for predicting the pressure rise time in areas with gas.

[0108] The predicted peak overpressure value for the open explosion zone (B1 zone) The output layer weights are the overpressure peak prediction model for the open explosion zone. The activation function for the overpressure peak prediction model in the open explosion zone is the Sigmoid function. The hidden layer weights are used in the overpressure peak prediction model for the open explosion zone. The characteristics of the fusion of the open zone of the explosion, This represents the hidden layer bias term in the overpressure peak prediction model for the open explosion zone. This is the output layer bias term of the overpressure peak prediction model for the open explosion zone; A mapping function for predicting the pressurization time in the open explosion zone;

[0109] Predicted pressure rise time for measuring points in the explosion reflection zone; Features of the fusion of the explosion reflection zone; , These are the hidden layer weights and bias terms of the explosion reflection zone pressurization time prediction model, respectively. , These are the output layer weights and bias terms of the explosion reflection zone pressurization time prediction model; The activation function for the explosion reflection zone pressure rise time prediction model is preferably the same as that used in the explosion reflection zone pressure rise time prediction model. This is a mapping function for predicting the pressure rise time in the explosion reflection zone.

[0110] For boost time prediction mapping function In areas with gas, the pressure rise time prediction mapping function Further combining gas concentration parameters and explosion source energy parameters to reflect the impact of gas participation in combustion and changes in explosion source conditions on pressurization time; in the explosion open zone (zone B1) and explosion reflection zone (zone B2), since there is no gas at the measuring point, the pressurization time prediction mapping function... Without introducing parameters such as gas concentration and explosion source energy, the boost time prediction mapping function is used in a specific application scenario. All use the Softplus function form, specifically expressed as follows:

[0111]

[0112] Therefore, by using the boost time prediction mapping function, the predicted boost time value is always positive while the prediction result changes smoothly with the input features, so as to be suitable for continuous regression prediction of boost time.

[0113] In this embodiment, there is a gas-fired zone ( The pressure rise in the explosion open zone (B1 zone) is mainly influenced by factors such as gas concentration, gas distribution range, and relative position of ignition source. The pressure rise in the explosion open zone (B2 zone) is mainly determined by the incident shock wave, showing obvious distance attenuation and saturation trends. The pressure rise time prediction model of the explosion open zone (B1 zone) fully and accurately reflects the changes in pressure rise time under different open conditions. The explosion reflection zone (B2 zone) has a complex structure, and the pressure-time exhibits a double-peak structure of incident peak and reflection peak. The pressure rise process in this area is not only affected by the incident wave, but also by the enhancement effect of structural reflection. The pressure rise time prediction model of the explosion reflection zone (B2 zone) automatically learns the rise law of incident peak and reflection peak, thereby accurately predicting the pressure rise time of the reflection zone.

[0114] Through the gas-filled area ( The pressure rise time prediction for the gas cloud zone can accurately characterize the time process of pressure rising from static pressure to peak value within the gas cloud zone by integrating the internal spatial characteristics of the gas cloud, the gas volume fraction, and the conditions of the explosion source, providing a reliable basis for risk level classification. The pressure rise time prediction for the explosion open zone (Zone B1) can effectively predict the pressure rise process of the explosion open zone (Zone B1) by comprehensively considering the location of the measuring point, the boundary characteristics of the gas cloud, and the openness conditions of the corridor. The pressure rise time prediction for the explosion reflection zone (Zone B2) can make full use of the structural reflection characteristics and bimodal dynamic information in the pressure sequence of the reflection zone, so that the pressure rise process of the reflection zone can be accurately predicted, thereby supporting the risk level determination.

[0115] In this embodiment, respectively in the gas-filled area ( Independent pressure rise time prediction models were constructed for the explosion open zone (B1 zone) and explosion reflection zone (B2 zone). Targeted analyses were conducted on the main influencing factors and pressure propagation mechanisms of the pressure rise process in different accident zones. The predictions yielded the time prediction results for gas pipeline leaks and explosions within the integrated utility tunnel. The overpressure peak value and pressure rise time at each measuring point are used to assess the risk level of an explosion accident, enabling zoned and refined risk early warning for gas pipeline explosion accidents. The risk assessment includes overpressure peak value and pressure rise time, and the destructiveness of vulnerable points within the integrated pipeline corridor is assessed based on established damage criteria. See Table 1 below:

[0116] Table 1 Damage Assessment of Integrated Utility Tunnels Based on Damage Criteria

[0117]

[0118] Based on the different objects of damage, the management level with the greatest damage to the affected object is selected as the comprehensive risk level, and a corresponding risk level response is output, including: minor risk, general risk, high risk, and major risk. Among these, gas leakage devices affect the gas leakage diffusion process; different gas leakage times correspond to different gas distribution ranges. Dynamic prediction of explosion overpressure damage risk can be integrated, and based on different measurement point distributions, a dynamic evolution of explosion risk for localized areas can be formed, guiding the central control platform operators of the utility tunnel company in rapid emergency response after a gas leakage accident. Furthermore, risk level visualization can be achieved through color-coded comparisons, and different risk warning forms can be provided based on different risks.

[0119] This implementation addresses gas leaks in confined spaces such as utility tunnels and tunnels by conducting a comprehensive risk zoning assessment of gas leak explosion accidents in utility tunnels. It establishes a dynamic zoning assessment model for gas explosion risks in utility tunnels, classifying and grading dynamic explosion overpressure damage. This effectively solves the problems of rapid extraction of explosion characteristics and efficient assessment of explosion risks in narrow, confined underground spaces, providing technical support for the safe operation of urban underground utility tunnels and the safe application of gas.

[0120] This embodiment also provides a risk classification and early warning system for gas pipeline leaks and explosions in integrated utility tunnels. It uses the risk classification and early warning method for gas pipeline leaks and explosions in integrated utility tunnels from any of the above embodiments to predict such leaks and explosions. Figure 8 , Figure 9 As shown, the system includes:

[0121] The simulation and data acquisition unit 901 is configured to simulate gas pipeline leaks and explosions in the integrated utility tunnel using an accident simulation device, and divide the integrated utility tunnel into different accident zones based on the simulation results; simultaneously, it collects data during the accident simulation process. Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing area identifiers for different accident zones are obtained; among them, It is a positive integer;

[0122] Leakage prediction unit 902 is configured to perform multi-feature fusion on the feature vectors containing area identifiers of different accident zones to obtain the leakage prediction result of a gas pipeline leak and explosion in the integrated utility tunnel. The overpressure peak value and pressure rise time at each measuring point.

[0123] The integrated utility tunnel gas pipeline leakage and explosion accident risk classification and early warning system provided in this embodiment can realize the steps and processes of any of the above-mentioned integrated utility tunnel gas pipeline leakage and explosion accident risk classification and early warning methods, and achieve the same technical effect, which will not be described in detail here.

[0124] In the description of this invention, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0125] In this invention, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0126] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A risk classification and early warning method for gas pipeline leakage and explosion accidents in integrated utility tunnels, characterized in that, include: Accident simulations of gas pipeline leaks and explosions in integrated utility tunnels were conducted using an accident simulation device. Based on the simulation results, the integrated utility tunnels were divided into different accident zones. The leakage volume rate obtained through the accident simulation was also analyzed. and duration of leakage According to the gas leak diffusion model: During a gas pipeline leak and explosion, the integrated utility tunnel is divided into a gas-bearing zone and a non-gas-bearing zone; where, This is the distance from the gas leak point to the boundary of the area containing the gas. The gas diffusion coefficient during a gas pipeline leak and explosion. The correction factor for the gas diffusion range under the ventilation conditions of the integrated utility tunnel; the non-gas zone is divided into the explosion reflection zone and the explosion open zone by the spatial parameters of the weak surface layout; Meanwhile, during the accident simulation process Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing area identifiers for different accident zones are obtained; among them, It is a positive integer; By fusing feature vectors of the included area identifiers from different accident zones, the results of gas pipeline leaks and explosions in integrated utility tunnels can be obtained. The overpressure peak value and pressure rise time at each measuring point.

2. The method according to claim 1, characterized in that, Based on the fusion model of gas-bearing area characteristics: Feature fusion is performed on feature vectors containing area identifiers of gas-fired zones; In the formula, For input characteristics of areas with gas supply, This is the feature fusion weight matrix for areas with gas supply; This is a spatial feature vector representing the distance from the measuring point in the gas-bearing area to the gas boundary, including the straight-line distance from the measuring point to the gas boundary. ; Features of the explosion source, including the energy characteristics of the explosion source. Explosion source area characteristics ; Characteristics of gas, For regional identification features, For the time-series pressure at the measuring point, This is the feature encoding bias term for the gas-zone feature fusion model. The spatial attenuation factor for areas with gas combustion; It has the characteristics of a gas-fired zone; The energy density of the explosion source.

3. The method according to claim 2, characterized in that, According to the overpressure peak prediction model for gas-filled areas: Predict the peak overpressure and pressure rise time in areas with gas combustion; In the formula, For the predicted overpressure peak in the gas-bearing area, The output layer weights of the overpressure peak prediction model for gas-filled areas are... Let be the activation function of the overpressure peak prediction model for gas-filled areas. The hidden layer weights are for the overpressure peak prediction model in the gas-filled zone. It has the characteristics of a gas-fired zone. This is the hidden layer bias term of the overpressure peak prediction model for gas-filled areas. This is the output layer bias term of the overpressure peak prediction model for gas-filled areas, where C represents the gas concentration. This is a correction term for gas concentration. This is a correction term for the energy source of the explosion. This is a characteristic of the explosion source energy.

4. The method according to claim 1, characterized in that, According to the fusion model of features of the open zone of the explosion: Feature fusion is performed on the feature vectors of the open blast zone that include area identifiers; In the formula, The characteristics of the fusion of the open zone of the explosion; The feature fusion weight matrix for the open zone of the explosion. This is the spatial feature vector of the distance from the measuring point in the explosion open zone to the gas boundary; This refers to the overpressure characteristics at the gas boundary of the open explosion zone. Characteristics of gas, For regional identification features, The straight-line distance from the measuring point in the open explosion zone to the gas boundary; This is the distance attenuation factor for the open zone of the explosion.

5. The method according to claim 4, characterized in that, According to the peak overpressure prediction model for the open zone of an explosion: Predict the peak overpressure and pressurization time in the open explosion zone; In the formula, The predicted peak overpressure in the open zone of the explosion. The output layer weights are the overpressure peak prediction model for the open explosion zone. is the activation function for the overpressure peak prediction model in the open explosion zone. The hidden layer weights are used in the overpressure peak prediction model for the open explosion zone. The characteristics of the fusion of the open zone of the explosion, This represents the hidden layer bias term in the overpressure peak prediction model for the open explosion zone. This is the output layer bias term of the overpressure peak prediction model for the open explosion zone.

6. The method according to claim 1, characterized in that, Based on the fusion model of explosion reflection zone characteristics: Feature fusion is performed on the feature vectors containing area identifiers in the explosion reflection zone; In the formula, To employ a cyclic network structure for the timing pressure of the explosion reflection zone Bimodal dynamic features obtained by sequence feature extraction; The time interval between the two pressure peaks in the time-series pressure sequence of the explosion reflection zone; The fusion characteristics of the explosion reflection zone The characteristic combination weights of the explosion reflection zone, This is the spatial feature vector of the distance from the measuring point in the explosion reflection zone to the gas boundary. This refers to the overpressure characteristics at the gas boundary of the explosion reflection zone. Characteristics of gas, For regional identification features, The distance is the straight-line distance from the measuring point in the explosion reflection zone to the gas boundary. The distance is the straight-line distance from the measuring point in the explosion reflection zone to the normally closed fire door inside the integrated utility tunnel. This represents the straight-line distance from the measuring point in the explosion reflection zone to the turning point inside the integrated utility tunnel. This is the straight-line distance from the measuring point in the explosion reflection zone to the branching point inside the integrated utility tunnel.

7. The method according to claim 6, characterized in that, According to the overpressure peak prediction model of the explosion reflection zone: Predicting the peak overpressure in the explosion reflection zone; In the formula, These are the predicted peak values ​​of incident pressure and reflected pressure in the explosion reflection zone, respectively. The output layer weights are the overpressure peak prediction model for the explosion reflection zone. is the activation function for the overpressure peak prediction model in the explosion reflection zone. The hidden layer weights are used in the overpressure peak prediction model for the explosion reflection zone. The fusion characteristics of the overpressure peak prediction model for the explosion reflection zone are used. This represents the hidden layer bias term in the overpressure peak prediction model for the explosion reflection zone. This is the output layer bias term of the overpressure peak prediction model for the explosion reflection zone; The distance is the straight-line distance from the measuring point in the explosion reflection zone to the normally closed fire door inside the integrated utility tunnel. This represents the straight-line distance from the measuring point in the explosion reflection zone to the turning point inside the integrated utility tunnel. The straight-line distance from the measuring point in the explosion reflection zone to the bifurcation point inside the integrated utility tunnel; The predicted overpressure peak value for the explosion reflection zone.

8. The method according to claim 1, characterized in that, According to the boost time prediction model: Predict the boost time for different accident zones; In the formula, For the predicted pressure rise time in areas with gas supply, These are the output layer weights and bias terms of the gas-filled zone pressure rise time prediction model, respectively. The activation function for the gas-fired zone pressure rise time prediction model is... These are the hidden layer weights and bias terms of the gas-filled zone pressure rise time prediction model; This represents the fusion characteristics of the gas-bearing zone; C represents the gas concentration. This is a correction term for gas concentration. This is a correction term for the energy of the explosion source; The predicted pressurization time for the open blast zone. These are the output layer weights and bias terms of the explosion open zone pressurization time prediction model; These are the hidden layer weights and bias terms of the explosion open zone pressurization time prediction model, respectively; The characteristics of the fusion of the open zone of the explosion; The activation function for the explosion open zone pressurization time prediction model; Predicted pressure rise time for measuring points in the explosion reflection zone; Features of the fusion of the explosion reflection zone; , These are the hidden layer weights and bias terms of the explosion reflection zone pressurization time prediction model, respectively. , These are the output layer weights and bias terms of the explosion reflection zone pressurization time prediction model; is the activation function for the explosion reflection zone pressurization time prediction model.

9. A risk classification and early warning system for gas pipeline leakage and explosion accidents in integrated utility tunnels, characterized in that, The integrated utility tunnel gas pipeline leakage and explosion accident risk classification and early warning method according to any one of claims 1-8 is used to predict the gas pipeline leakage and explosion in the integrated utility tunnel. The system includes: The simulation and data acquisition unit is configured to simulate gas pipeline leaks and explosions in the integrated utility tunnel using an accident simulation device, and divide the integrated utility tunnel into different accident zones based on the simulation results; among them, the leakage volume rate obtained through the accident simulation is... and duration of leakage According to the gas leak diffusion model: During a gas pipeline leak and explosion, the integrated utility tunnel is divided into a gas-bearing zone and a non-gas-bearing zone; where, This is the distance from the gas leak point to the boundary of the area containing the gas. The gas diffusion coefficient during a gas pipeline leak and explosion. The correction factor for the gas diffusion range under the ventilation conditions of the integrated utility tunnel; the non-gas zone is divided into the explosion reflection zone and the explosion open zone by the spatial parameters of the weak surface layout; Meanwhile, during the accident simulation process Based on the spatial parameters and temporal pressure of each measuring point, feature vectors containing area identifiers for different accident zones are obtained; among them, It is a positive integer; The leak prediction unit is configured to perform multi-feature fusion on the feature vectors containing area identifiers of different accident zones to obtain the leak prediction result of a gas pipeline leak and explosion in the integrated utility tunnel. The overpressure peak value and pressure rise time at each measuring point.

Images