CN-121545298-B - Comprehensive pipe rack gas pipeline leakage explosion accident risk grading early warning method and system

CN121545298BCN 121545298 BCN121545298 BCN 121545298BCN-121545298-B

Abstract

The application relates to the technical field of urban underground engineering, and provides a comprehensive pipe rack gas pipeline leakage explosion accident risk grading early warning method and system. Accident deduction is carried out on leakage and explosion of a gas pipeline of the comprehensive pipe rack, the comprehensive pipe rack is divided into different accident partitions according to deduction results, and meanwhile, the accident deduction process is collected The space parameters and time sequence pressure of each measuring point are used for obtaining the characteristic vectors containing the area identifiers of different accident partitions, and then the characteristic vectors containing the area identifiers of different accident partitions are subjected to multi-characteristic fusion to obtain the comprehensive pipe gallery when the gas pipeline leaks and explodes And the overpressure peak value and the pressure rise time of each measuring point are used for realizing comprehensive risk partition assessment of the comprehensive pipe rack leakage explosion accident, forming a partition dynamic assessment mode of the comprehensive pipe rack gas explosion risk, and carrying out partition grading on the dynamic explosion overpressure induced damage, thereby effectively solving the problems of rapid extraction of the gas leakage explosion characteristics and efficient assessment of the explosion risk of the underground long and narrow limited space.

Inventors

WU JIANSONG
CAO JIAOJIAO
WANG YUBO
XU TONG

Assignees

中国矿业大学(北京)

Dates

Publication Date: 20260505
Application Date: 20260116

Claims (9)

1. The utility tunnel gas pipeline leakage explosion accident risk grading early warning method is characterized by comprising the following steps of: Accident deduction is carried out on leakage and explosion of the gas pipeline of the comprehensive pipe gallery through an accident deduction device, and the comprehensive pipe gallery is divided into different accident areas according to deduction results, wherein the leakage volume rate is obtained through the accident deduction And leak duration According to a gas leakage diffusion model: in the leakage and explosion process of the gas pipeline, the utility tunnel is divided into a gas area and a gas-free area, The distance from the gas leakage point to the boundary of the gas area is the distance; is the gas diffusion coefficient in the gas pipeline leakage explosion process, Dividing a gas-free area into an explosion reflection area and an explosion opening area through the space parameters of the weak surface layout; meanwhile, in the process of collecting accident deduction The space parameters and time sequence pressure of each measuring point are used for obtaining the characteristic vectors containing the area identifiers of different accident partitions, Is a positive integer; The characteristic vectors containing the area identifiers of different accident partitions are subjected to multi-characteristic fusion to obtain the comprehensive pipe rack when the gas pipeline leaks and explodes Overpressure peak and boost time at each station.
2. The method of claim 1, wherein the gas zone feature fusion model is defined as: Feature fusion is carried out on feature vectors containing area identifiers in the gas area; In the formula, For the input features of the gas zone, Fusing a weight matrix for the characteristics of the gas area; For the space feature vector with the distance from the measuring point of the gas area to the gas boundary, the space feature vector comprises the linear distance from the measuring point to the gas boundary ; Is characteristic of explosive source, including characteristic of energy of explosive source Area characteristics of explosion source ; As a feature of the gas it is, A feature is identified for the region and, As the time-series pressure of the measuring point, Bias items are encoded for the features of the feature fusion model of the gas area, Is the space attenuation factor of the gas area; is a fusion feature of the gas zone; Is the energy density of the explosion source.
3. The method of claim 2, wherein the gas area overpressure peak prediction model is: predicting overpressure peak value and boosting time of a gas area; In the formula, For a predicted overpressure peak in a gas zone, For the output layer weight of the gas area overpressure peak prediction model, For an activation function of the gas area overpressure peak prediction model, For hidden layer weights of the gas area overpressure peak prediction model, In order to have the merging feature of the gas zones, For the hidden layer bias term of the gas area overpressure peak prediction model, The bias term of the output layer of the overpressure peak prediction model of the gas area is provided, C is the gas concentration, Is a gas concentration correction term; As an energy correction term of the explosion source, Is the energy characteristic of the explosion source.
4. The method of claim 1, wherein the model is fused according to an explosion open area feature: feature fusion is carried out on feature vectors containing area identifiers in the explosion open area; In the formula, Fusion features that are explosive open areas; the weight matrix is fused for the characteristics of the explosion open area, A space feature vector which is the distance from the measuring point of the explosion open area to the gas boundary; as an overpressure feature at the gas boundary of the explosion open area, As a feature of the gas it is, A feature is identified for the region and, The linear distance from the measuring point of the explosion open area to the gas boundary; is the distance attenuation factor of the explosive open area.
5. The method of claim 4, wherein the explosion open area overpressure peak prediction model is based on: predicting overpressure peak value and boosting time of an explosion open area; In the formula, For the predicted overpressure peak of the explosive open area, The output layer weight of the explosion open area overpressure peak prediction model, As an activation function of the explosion open area overpressure peak prediction model, The hidden layer weight of the explosion open area overpressure peak prediction model, As a fusion feature of the explosive open area, A hidden layer bias term for an explosion open area overpressure peak prediction model, And (5) outputting layer bias items for the explosion open area overpressure peak prediction model.
6. The method of claim 1, wherein the model is fused according to an explosive reflection zone feature: Feature fusion is carried out on feature vectors containing region identifiers in the explosion reflection region; In the formula, To adopt a circulating network structure to carry out time sequence pressure on an explosion reflection area Extracting sequence features to obtain bimodal dynamic features; time intervals of double pressure peaks in the time sequence pressure sequence of the explosion reflection area; As a fusion feature of the explosive reflection area, The weights are combined for the features of the explosive reflection zone, The space characteristic vector of the distance from the measuring point of the explosion reflection area to the gas boundary, As an overpressure feature at the gas boundary of the explosive reflection zone, As a feature of the gas it is, A feature is identified for the region and, For the straight line distance from the explosive reflection area measuring point to the gas boundary, For the straight line distance from the measuring point of the explosion reflection area to the normally closed fireproof door in the comprehensive pipe rack, For the straight line distance from the measuring point of the explosion reflection area to the turning point in the utility tunnel, The straight line distance from the measuring point of the explosion reflection area to the bifurcation point in the utility tunnel.
7. The method of claim 6, wherein the peak overpressure prediction model for the explosive reflection zone is based on: Predicting an overpressure peak value of the explosion reflection area; In the formula, The method comprises the steps of respectively predicting a peak value of incident pressure and a peak value of reflected pressure of an explosion reflection area; the output layer weight of the overpressure peak prediction model of the explosion reflection area, As an activation function of the blast reflection area overpressure peak prediction model, The hidden layer weight of the explosion reflection area overpressure peak prediction model, As a fusion feature of the explosion reflection area overpressure peak prediction model, A hidden layer bias term for the explosive reflection area overpressure peak prediction model, An output layer bias term of an explosion reflection area overpressure peak value prediction model; for the straight line distance from the measuring point of the explosion reflection area to the normally closed fireproof door in the comprehensive pipe rack, For the straight line distance from the measuring point of the explosion reflection area to the turning point in the utility tunnel, The straight line distance from the measuring point of the explosion reflection area to the bifurcation point in the comprehensive pipe rack; a predicted overpressure peak for the explosive reflection zone.
8. The method of claim 1, wherein the step-up time prediction model is: Predicting the boosting time of different accident partitions; In the formula, For a predicted boost time for a gas zone, Respectively outputting layer weights and paranoid items of the boost time prediction model of the gas area; For the activation function of the gas area boost time prediction model, Respectively a hidden layer weight and a bias term of a boost time prediction model of the gas area; Is the fusion characteristic of the gas area, C is the concentration of the gas, Is a gas concentration correction term; is an explosion source energy correction term; For the predicted boost time of the explosive open area, Respectively outputting layer weights and bias items of the explosion open area boost time prediction model; respectively obtaining hidden layer weight and bias items of a boost time prediction model of an explosion open area; Fusion features that are explosive open areas; An activation function of a boost time prediction model for the explosion open area; The predicted boosting time of the measuring point of the explosion reflection area; Fusing features for the explosive reflection zone; 、 respectively obtaining hidden layer weight and bias items of the explosion reflection area boosting time prediction model; 、 respectively outputting layer weights and bias items of the explosion reflection area boosting time prediction model; And (3) an activation function of the boost time prediction model for the explosion reflection area.
9. A utility tunnel gas pipeline leakage explosion accident risk classification early warning system, characterized in that the utility tunnel gas pipeline leakage explosion accident risk classification early warning method according to any one of claims 1-8 is adopted to predict utility tunnel gas pipeline leakage explosion, the system comprises: The deduction acquisition unit is configured to deduct accidents of the leakage explosion of the gas pipeline of the comprehensive pipe gallery through the accident deduction device, and divide the comprehensive pipe gallery into different accident partitions according to deduction results, wherein the leakage volume rate is obtained through accident deduction And leak duration According to a gas leakage diffusion model: in the leakage and explosion process of the gas pipeline, the utility tunnel is divided into a gas area and a gas-free area, The distance from the gas leakage point to the boundary of the gas area is the distance; is the gas diffusion coefficient in the gas pipeline leakage explosion process, Dividing a gas-free area into an explosion reflection area and an explosion opening area through the space parameters of the weak surface layout; meanwhile, in the process of collecting accident deduction The space parameters and time sequence pressure of each measuring point are used for obtaining the characteristic vectors containing the area identifiers of different accident partitions, Is a positive integer; The leakage prediction unit is configured to perform multi-feature fusion on the feature vectors containing the region identifiers of different accident partitions to obtain the leakage explosion time of the gas pipeline of the comprehensive pipe rack Overpressure peak and boost time at each station.

Description

Comprehensive pipe rack gas pipeline leakage explosion accident risk grading early warning method and system Technical Field The application relates to the technical field of urban underground engineering, in particular to a comprehensive pipe rack gas pipeline leakage explosion accident risk grading early warning method and system. Background In infrastructure such as urban underground utility tunnel, the risk of pipeline leakage explosion accident is very big, and traditional gas monitoring early warning has apparent technical bottleneck. The gas pipe provides a gas source for a gas explosion accident, and the easy-to-fire feature of the high-voltage power line provides an explosion source for the gas explosion accident, however, the long and narrow underground tunnel easily accumulates a large amount of gas and rapidly spreads to expand the gas-related range. Meanwhile, the existence of a large number of high-risk urban life lines (fuel gas, electric power, heating ventilation, water supply and drainage and the like) in a long and narrow urban underground space makes fuel gas explosion more complex, for example, electric fire and high-voltage explosion are easily caused by an electric power pipeline, and a high-temperature surface is caused by a heating pipeline. In addition, open fire operation, unexpected static electricity and the like of operators in urban underground spaces become accidental fire sources. In an underground long and narrow confined space of up to several tens of kilometers, explosion energy generated after explosion occurs is not easy to release, and huge energy accumulation is caused, so that the consequences are extremely dangerous. Disclosure of Invention The application aims to provide a comprehensive pipe rack gas pipeline leakage accident risk grading early warning method and system, which are used for solving or relieving the problems in the prior art. In order to achieve the above object, the present application provides the following technical solutions: The application provides a comprehensive pipe rack gas pipeline leakage explosion accident risk grading early warning method, which comprises the following steps: carrying out accident deduction on leakage and explosion of the gas pipeline of the comprehensive pipe gallery through an accident deduction device, dividing the comprehensive pipe gallery into different accident areas according to deduction results, and collecting the accident deduction process at the same time The space parameters and time sequence pressure of each measuring point are used for obtaining the characteristic vectors containing the area identifiers of different accident partitions,Is a positive integer; The characteristic vectors containing the area identifiers of different accident partitions are subjected to multi-characteristic fusion to obtain the comprehensive pipe rack when the gas pipeline leaks and explodes Overpressure peak and boost time at each station. Preferably, the leak volume rate obtained by accident deductionAnd leak durationAccording to a gas leakage diffusion model: in the leakage and explosion process of the gas pipeline, the utility tunnel is divided into a gas area and a gas-free area, The distance from the gas leakage point to the boundary of the gas area is the distance; is the gas diffusion coefficient in the gas pipeline leakage explosion process, The correction coefficient of the gas diffusion range under the ventilation condition of the comprehensive pipe rack is obtained; The gas-free area is divided into an explosion reflection area and an explosion opening area by the space parameters of the weak surface layout. Preferably, according to the feature fusion model of the gas area: Feature fusion is carried out on feature vectors containing area identifiers in the gas area; In the formula, For the input features of the gas zone,Fusing a weight matrix for the characteristics of the gas area; For the space feature vector with the distance from the measuring point of the gas area to the gas boundary, the space feature vector comprises the linear distance from the measuring point to the gas boundary ;Is characteristic of explosive source, including characteristic of energy of explosive sourceArea characteristics of explosion source;As a feature of the gas it is,A feature is identified for the region and,As the time-series pressure of the measuring point, Bias items are encoded for the features of the feature fusion model of the gas area,Is the space attenuation factor of the gas area; is a fusion feature of the gas zone; Is the energy density of the explosion source. Preferably, according to the gas area overpressure peak prediction model: predicting an overpressure peak value of a gas area; In the formula, For a predicted overpressure peak in a gas zone,For the output layer weight of the gas area overpressure peak prediction model,For an activation function of the gas area overpressure peak prediction model,For hidden layer