Highway tunnel emergency evacuation path planning method, system, equipment and medium

Abstract

The invention provides a highway tunnel emergency evacuation path planning method, system, device and medium, and the method comprises the steps: collecting disaster data at a fixed time interval, generating a plurality of selectable escape paths according to the disaster data, and calculating the real escape rates of all selectable escape paths; calculating a predicted escape rate of each optional escape path according to the disaster data; dynamically generating a correction weight according to the change degree of the poison gas concentration at the current moment and the last moment; and dynamically correcting the real escape rate of each optional path according to the correction weight, and selecting the optional escape path with the highest real escape rate as an evacuation path. According to the method and the device, the problems that in the prior art, the planning of the escape path depends on a fixed topology model and a preset rule, and only the shortest path is taken as an optimization target, so that the generated escape path is disjointed with disaster evolution, and the planned path does not accord with the actual condition are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of emergency response in highway tunnels, and in particular to a method, system, equipment and medium for planning an emergency evacuation path in a highway tunnel. Background Art

[0002] As highway tunnels expand and safety standards improve, traditional emergency evacuation systems face increasingly prominent challenges in path planning and equipment coordination in complex disaster scenarios. Existing technologies generally employ static path generation models, relying on fixed topological models and preset rules. These technologies are unable to integrate dynamic variables such as fire source location, toxic gas diffusion, and equipment status in real time. This leads to a disconnect between escape routes and disaster evolution, creating safety hazards where the planned path does not match the actual situation. Furthermore, traditional path planning algorithms fail to fully consider the dynamic impact of environmental parameters such as toxic gas diffusion and thermal radiation on escape routes. They focus solely on the shortest path and ignore the balancing of multiple safety objectives, making it difficult to generate an optimal solution that balances safety and timeliness in emergency situations. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides a method for planning emergency evacuation paths in highway tunnels, which solves the problem that the planning of escape paths in the existing technology relies on fixed topological models and preset rules, and only takes the shortest path as the optimization goal, resulting in the generated escape path being out of touch with the evolution of disasters, and the planned path not being consistent with the actual situation.

[0004] According to an embodiment of the present invention, a method for planning an emergency evacuation path in a highway tunnel includes:

[0005] Collect disaster data at fixed time intervals, generate multiple optional escape routes based on the disaster data, and calculate the actual escape rates of all optional escape routes;

[0006] Calculate the predicted escape rate for each optional escape route based on the hazard data;

[0007] Dynamically generate correction weights based on the degree of change in poison gas concentration between the current moment and the previous moment;

[0008] According to the correction weight, the true escape rate of each optional path is dynamically corrected, and the optional escape path with the highest true escape rate is selected as the evacuation path.

[0009] Preferably, the method for calculating the predicted escape rate of each optional escape path based on disaster data includes:

[0010] Collect scene topology information of highway tunnels, build a digital mirror grid of the tunnel environment based on the scene topology information, and generate multiple virtual pedestrians;

[0011] Use all virtual pedestrians to conduct escape simulation for each optional escape route, and use the toxic gas diffusion model to calculate the cumulative exposure index of each virtual pedestrian;

[0012] The ratio of the number of virtual pedestrians whose cumulative exposure index does not exceed the threshold and who successfully escape to the total number of virtual pedestrians is used as the predicted escape rate of the corresponding optional escape path.

[0013] Preferably, the calculation formula of the cumulative exposure index is as follows:

[0014]

[0015] Where C(x,y,z,t) is the real-time gas concentration of the grid with coordinates (x,y,x) in the digital mirror grid of the tunnel environment at time t, Δt is the time step (5 seconds), is the toxicological absorption coefficient of the grid with coordinates (x, y, x) in the digital mirror grid of the tunnel environment, v i (t) The real-time speed of the i-th virtual pedestrian, and T is the total time.

[0016] Preferably, when constructing the digital mirror grid of the tunnel environment, passages where the concentration of toxic gases exceeds a threshold and has no ventilation capacity need to be eliminated.

[0017] Preferably, the correction weight includes a time weight α and a toxic gas exposure weight β;

[0018] The correction formula of the time weight α is as follows

[0019]

[0020] The correction formula for the toxic gas exposure weight β is:

[0021]

[0022] Where Q is the exposure amount of toxic gas, s t is the gas concentration at the current moment, s t-1 is the gas concentration at the previous moment, μ0 is the basic coefficient, R actual is the true escape rate, R predicted To predict the escape rate, t is the time and η is the learning rate.

[0023] Preferably, the sum of the time weight and the toxic gas exposure weight cannot be greater than 1.2.

[0024] Preferably, after obtaining the predicted escape rate, if the deviation between the predicted escape rate and the actual escape rate of the optional escape path is greater than 5%, the turbulent viscosity coefficient is calculated using the Navier-Stokes equation based on the predicted escape rate, and the model accuracy of the poison gas diffusion model is adjusted based on the turbulent viscosity coefficient, and then the predicted escape rate is recalculated;

[0025] Repeat this step until the deviation between the predicted escape rate and the actual escape rate is within 3%.

[0026] On the other hand, according to an embodiment of the present invention, a system for planning an emergency evacuation route in a highway tunnel is further provided. The system uses the above-mentioned method for planning an emergency evacuation route in a highway tunnel, including:

[0027] An edge module, which is used to collect disaster data and scene topology information in highway tunnels and construct a digital mirror grid of the tunnel environment based on the scene topology information;

[0028] A prediction module, configured to calculate a predicted escape rate for each optional escape path based on disaster data;

[0029] A correction module, the correction module is used to construct a poison gas diffusion model and generate correction weights using the poison gas diffusion model based on disaster data;

[0030] A decision module is used to calculate the actual escape rate of each optional escape path and determine the evacuation path according to the actual escape rate.

[0031] On the other hand, according to an embodiment of the present invention, a computer is also provided, comprising at least one processor and a memory, wherein the memory stores a computer program, and the computer program is configured to be executed by the processor to implement the above-mentioned method for planning an emergency evacuation path in a highway tunnel.

[0032] On the other hand, according to an embodiment of the present invention, a storage medium is also provided, which is a computer-readable storage medium. A computer program is stored on the storage medium, and the computer program can be executed by one or more processors to implement the above-mentioned highway tunnel emergency evacuation path planning method.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The present invention generates multiple optional escape paths based on the disaster data at the time of the disaster, and then generates virtual pedestrians to simulate escape based on the scene topology information of the highway tunnel, and predicts the predicted escape rate of each optional escape path. Since the concentration of toxic gas in the highway tunnel is constantly changing after the disaster occurs, the disaster data collected each time is also constantly changing, so the generated optional escape path will also be constantly changing. Then, correction weights are generated according to the toxic gas diffusion model. As the toxic gas concentration in the highway tunnel continues to change, the correction weights generated by the toxic gas diffusion model are also constantly changing. Therefore, as time goes by, the actual escape rate of the generated optional escape path is also constantly changing. Therefore, during the escape process of trapped people, the escape path is also constantly changing according to the disaster environment, and no longer relies on fixed topology models and preset rules. At the same time, since the actual escape rate is mainly affected by the toxic gas concentration, the shortest path is no longer the optimization goal. The generated escape path is closely related to the evolution of the disaster. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flowchart of emergency evacuation route planning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0037] like Figure 1 As shown, an embodiment of the present invention proposes a method for planning an emergency evacuation path in a highway tunnel, comprising:

[0038] Collect disaster data at fixed time intervals, generate multiple optional escape routes based on the disaster data, and calculate the actual escape rates of all optional escape routes;

[0039] Various sensors deployed at highway tunnel sites collect on-site disaster data, and quickly and locally process tunnel electromechanical equipment operating status data, video surveillance data, and ambient temperature data. Association rule analysis is used to mine cross-modal correlation features such as "sudden CO concentration increase and video fire point coordinates." Ultimately, a comprehensive tunnel situation map is generated, marking the fire source location, toxic gas diffusion range and concentration, structural damage areas, and a heat map of trapped personnel distribution. This provides a panoramic data foundation for subsequent decision-making, avoiding the global path planning lag caused by data transmission delays in traditional centralized architectures, and ensuring that the optimal escape route that fits the local environment can be generated at the early stages of a disaster.

[0040] After collecting disaster data, we use the existing path planning algorithm to generate multiple optional escape paths, and calculate the actual escape rate of each optional escape path according to the following formula:

[0041] R actual =α*T+β*Q+γ*Z

[0042] Where T is the estimated time it takes for the escapee to reach the exit from their current location; Q is the gas exposure, which is the cumulative gas concentration inhaled by the escapee along the escape path (ppm·s); Z is the path complexity, which is the number of turns, the proportion of steep slopes, and the number of narrow passages in the escape path; α is the time weighting coefficient; β is the gas exposure weighting coefficient; and γ is a constant, which is set to 0.2.

[0043] Since the environment in the highway tunnel is constantly changing, the weight of the escape time and the weight of the toxic gas exposure should also change accordingly. Therefore, the optimization goal will not be just the shortest path. The dynamic weight coefficients α and β are the key parameters for adaptively adjusting the priority of the multi-objective optimization. α controls the weight of the escape time in the objective function, with an initial value of 0.5. β is the weight coefficient for controlling the toxic gas exposure, with an initial value of 0. The adjustment of α and β is based on the actual escape rate (R actual ) and the predicted escape rate (R predicted ), when the gas concentration is low at the beginning of the fire, α dominates to shorten the escape time; when the gas concentration rises sharply, β increases dynamically to avoid high-risk areas. Through dynamic updates of the reinforcement learning formula, a composite path plan is generated that takes into account the shortest distance, low-risk areas, and emergency resource distribution, breaking through the mechanical and hysteresis of traditional static path planning.

[0044] Calculate the predicted escape rate for each optional escape route based on the hazard data;

[0045] The generation of the predicted escape rate relies on a three-dimensional space-time grid model and a dynamic scenario deduction engine. The system integrates the gas concentration gradient tensor (0.5m 3 A digital mirror of the tunnel environment is constructed using multi-source data, including high-resolution data, occupant distribution heat maps, ventilation efficiency coefficients, and dynamic obstacle topology. Within the discrete event simulation framework, passages with excessive toxic gas concentrations and no ventilation capacity (or low ventilation efficiency) are first eliminated. 2,000 virtual pedestrians are then deployed, including 1,000 young people moving at a speed of 2.0 m / s ± 0.5 and 1,000 elderly people moving at a speed of 1.2 m / s ± 0.3. All virtual pedestrians are then used to simulate escape attempts for each optional escape path and calculate the cumulative exposure index. Escape is considered a failure when the CEI exceeds the toxicological lethal threshold. This cumulative exposure index is calculated using the spatiotemporal integral formula:

[0046]

[0047] Where C(x,y,z,t) is the real-time gas concentration of the grid with coordinates (x,y,x) in the digital mirror grid of the tunnel environment at time t, Δt is the time step (5 seconds), which represents the interval between each data update. is the toxicological absorption coefficient of the grid with coordinates (x, y, x) in the digital mirror grid of the tunnel environment, which is related to the type of toxic gas and the individual breathing rate. i (t) The real-time speed of the i-th virtual pedestrian, which is affected by the speed distribution of virtual pedestrians and obstacles, and T is the total time.

[0048] Among the 2000 virtual pedestrians deployed, the ratio of the number of virtual pedestrians whose cumulative exposure index does not exceed the threshold and who escape successfully to the total number of virtual pedestrians is used as the predicted escape rate of the corresponding optional escape path.

[0049] In addition, if the deviation between the predicted escape rate and the actual escape rate of the optional escape path is greater than 5%, the turbulent viscosity coefficient is calculated based on the predicted escape rate using the Navier-Stokes equation:

[0050]

[0051] Among them, μ is the turbulent viscosity coefficient, C is the mass density of the poison gas, u is the flow velocity of the poison gas, and the adjustment range of μ value is 0.1~0.5m 2 / s, used to optimize the accuracy of the poison gas diffusion model (increasing μ makes the concentration distribution smoother, and decreasing μ enhances the local gradient).

[0052] The model accuracy of the poison gas diffusion model is then adjusted according to the turbulent viscosity coefficient, and the predicted escape rate is recalculated. This step is repeated until the deviation between the predicted escape rate and the actual escape rate is within 3%.

[0053] Dynamically generate correction weights based on the degree of change in poison gas concentration between the current moment and the previous moment;

[0054] The correction weights include a time weight α and a toxic gas exposure weight β;

[0055] The correction formula of the time weight α is as follows

[0056]

[0057] The correction formula for the toxic gas exposure weight β is:

[0058]

[0059] Where Q is the exposure amount of toxic gas, s t is the gas concentration at the current moment, s t-1 is the gas concentration at the previous moment, μ0 is the basic coefficient, R actual is the true escape rate, R predicted To predict the escape rate, t is the time, η (default 0.01) is the learning rate, which controls the update speed, and the partial derivative and Reflects the direction of impact of changes in α and β on the overall reward.

[0060] When the toxic gas concentration is low at the beginning of a fire, α dominates to shorten the escape time; when the toxic gas concentration rises sharply, β increases dynamically to avoid high-risk areas. At the same time, the toxic gas diffusion model is calibrated by μ to improve the accuracy of weight adjustment.

[0061] The toxic gas diffusion model dynamically adjusts the path planning weight parameters through the nonlinear mapping relationship between concentration and risk. When the toxic gas concentration increases from the initial 50 ppm to 300 ppm, the time weight α gradually decreases from 0.4 to 0.2 and the toxic gas exposure weight β increases from 0.4 to 0.6. These settings are based on the threshold effect of toxic exposure and multi-objective optimization constraints.

[0062] Experimental data shows that human tolerance to toxic gas concentrations exceeding 200 ppm decreases exponentially. Therefore, a staged adjustment strategy is adopted: in the low-risk phase (50-150 ppm), α is gradually reduced at a rate of 0.05 per 10 ppm, while also balancing traffic efficiency. In the high-risk phase (150-300 ppm), the reduction rate of α is increased to 0.1 per 10 ppm, while β is accelerated using a Sigmoid function to prevent weight oscillation and strengthen safety dominance. The upper limit of α + β of 1.2 is derived from multi-objective optimization Pareto frontier analysis: when α + β exceeds this threshold, the marginal benefits of path safety and timeliness are significantly reduced. For example, if β increases to 1.0, the path circuitry increases sharply, exceeding the toxicity tolerance time window; while an increase of β to 0.5 makes it impossible to effectively avoid high-risk areas. Through collaborative optimization using a toxic gas diffusion CFD model and reinforcement learning, it was determined that a β of 0.6 is sufficient for controlling the path detour distance and keeping the toxic gas exposure below 75% of the safety limit.

[0063] Finally, the corrected weights are used to correct the true escape rate of each optional path, and the optional escape path with the highest true escape rate is selected as the evacuation path. Since the concentration of toxic gas in the highway tunnel is constantly changing after the disaster, the disaster data collected each time is also constantly changing, so the generated optional escape path will also be constantly changing, and the corrected weights generated by the toxic gas diffusion model are also constantly changing. Therefore, with the passage of time, the true escape rate of the generated optional escape path is also constantly changing. Therefore, in the process of escaping trapped people, the escape path is also constantly changing according to the disaster environment, and no longer relies on fixed topological models and preset rules. At the same time, since the true escape rate is mainly affected by the toxic gas concentration, the shortest path is no longer the optimization goal, and the generated escape path is closely related to the evolution of the disaster.

[0064] On the other hand, an embodiment of the present invention further provides a system for planning emergency evacuation routes in highway tunnels. The system uses the above-mentioned method for planning emergency evacuation routes in highway tunnels, including:

[0065] An edge module, which is used to collect disaster data and scene topology information in highway tunnels and construct a digital mirror grid of the tunnel environment based on the scene topology information;

[0066] A prediction module, configured to calculate a predicted escape rate for each optional escape path based on disaster data;

[0067] A correction module, the correction module is used to construct a poison gas diffusion model and generate correction weights using the poison gas diffusion model based on disaster data;

[0068] A decision module is used to calculate the actual escape rate of each optional escape path and determine the evacuation path according to the actual escape rate.

[0069] On the other hand, an embodiment of the present invention further provides a computer comprising at least one processor and a memory, wherein the memory stores a computer program, and the computer program is configured to be executed by the processor to implement the above-mentioned method for planning an emergency evacuation path in a highway tunnel.

[0070] On the other hand, an embodiment of the present invention further provides a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and the computer program can be executed by one or more processors to implement the above-mentioned highway tunnel emergency evacuation path planning method.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for planning an emergency evacuation route in a highway tunnel, characterized by: include: Collect disaster data at fixed time intervals, generate multiple optional escape routes based on the disaster data, and calculate the actual escape rates of all optional escape routes; Calculate the predicted escape rate for each optional escape route based on the hazard data; Dynamically generate correction weights based on the degree of change in poison gas concentration between the current moment and the previous moment; According to the correction weight, the true escape rate of each optional path is dynamically corrected, and the optional escape path with the highest true escape rate is selected as the evacuation path.

2. A method for planning an emergency evacuation route in a highway tunnel according to claim 1, characterized in that: Methods for calculating the predicted escape rate for each optional escape route based on hazard data include: Collect scene topology information of highway tunnels, build a digital mirror grid of the tunnel environment based on the scene topology information, and generate multiple virtual pedestrians; Use all virtual pedestrians to conduct escape simulation for each optional escape route, and use the toxic gas diffusion model to calculate the cumulative exposure index of each virtual pedestrian; The ratio of the number of virtual pedestrians whose cumulative exposure index does not exceed the threshold and who successfully escape to the total number of virtual pedestrians is used as the predicted escape rate of the corresponding optional escape path.

3. The method for planning an emergency evacuation route in a highway tunnel according to claim 1, wherein: The calculation formula of the cumulative exposure index is as follows: Where C(x,y,z,t) is the real-time gas concentration of the grid with coordinates (x,y,x) in the digital mirror grid of the tunnel environment at time t, Δt is the time step (5 seconds), is the toxicological absorption coefficient of the grid with coordinates (x, y, x) in the digital mirror grid of the tunnel environment, v i (t) The real-time speed of the i-th virtual pedestrian, and T is the total time.

4. The method for planning an emergency evacuation route in a highway tunnel according to claim 1, wherein: When constructing the digital mirror grid of the tunnel environment, it is necessary to eliminate channels where the toxic gas concentration exceeds the threshold and has no ventilation capacity.

5. The method for planning an emergency evacuation route in a highway tunnel according to claim 1, wherein: The correction weights include a time weight α and a toxic gas exposure weight β; The correction formula of the time weight α is as follows The correction formula for the gas exposure weight β is: Where Q is the exposure amount of toxic gas, s t is the gas concentration at the current moment, s t-1 is the gas concentration at the previous moment, μ0 is the basic coefficient, R actual is the true escape rate, R predicted To predict the escape rate, t is the time and η is the learning rate.

6. The method for planning an emergency evacuation route in a highway tunnel according to claim 1, wherein: The sum of the time weight and the toxic gas exposure weight cannot be greater than 1.

2.

7. The method for planning an emergency evacuation route in a highway tunnel according to claim 1, wherein: After obtaining the predicted escape rate, if the deviation between the predicted escape rate and the actual escape rate for the optional escape path is greater than 5%, the turbulent viscosity coefficient is calculated using the Navier-Stokes equation based on the predicted escape rate, and the model accuracy of the poison gas diffusion model is adjusted based on the turbulent viscosity coefficient, and then the predicted escape rate is recalculated; Repeat this step until the deviation between the predicted escape rate and the actual escape rate is within 3%.

8. A system for planning emergency evacuation routes in highway tunnels, characterized in that: The system uses a highway tunnel emergency evacuation path planning method according to any one of claims 1 to 7, comprising: An edge module, which is used to collect disaster data and scene topology information in highway tunnels and construct a digital mirror grid of the tunnel environment based on the scene topology information; A prediction module, configured to calculate a predicted escape rate for each optional escape path based on disaster data; A correction module, the correction module is used to construct a poison gas diffusion model and generate correction weights using the poison gas diffusion model based on disaster data; A decision module is used to calculate the actual escape rate of each optional escape path and determine the evacuation path according to the actual escape rate.

9. A computer, characterized in that: The method comprises at least one processor and a memory, wherein the memory stores a computer program, and the computer program is configured to be executed by the processor to implement the method for planning an emergency evacuation path for a highway tunnel according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. The computer program can be executed by one or more processors to implement a highway tunnel emergency evacuation path planning method as described in any one of claims 1 to 7.

Images