A dynamic guiding sign optimization method

Abstract

The application relates to a dynamic guiding sign optimization method, which comprises the following steps: dividing the internal evacuation space of a subway station into multiple cells, and constructing a space discrete model according to the actual space structure, a preset channel connection relationship and a preset evacuation; defining the number of passengers in each cell and the passenger transmission flow between adjacent cells in the current time step, and deducing the evacuation process of the passengers in the multiple cells based on preset supply and demand constraints to construct a cell transmission model; mapping the fire environment parameters to the cell level to correct the traffic capacity of the cell in the fire environment; constructing a dynamic guiding sign optimization model, embedding the uncertainty in the fire environment into the dynamic guiding sign optimization model to form a dynamic guiding sign robust optimization model; and solving the robust optimization model and obtaining the optimal dynamic guiding sign pointing strategy through iterative updating. The intelligent level of evacuation guidance in the fire environment is significantly improved, and effective technical support is provided for the emergency management of large subway stations.

Description

Technical Field

[0001] This invention relates to the field of urban rail transit safety and emergency evacuation technology, and in particular to a method for optimizing dynamic guidance signs. Background Technology

[0002] With the rapid development of urban rail transit, subways have become an important part of urban transportation. However, subway stations are usually located underground, characterized by enclosed spaces, complex structures, and high population density. In the event of a fire, smoke spreads rapidly, visibility decreases sharply, and congestion can easily occur, seriously threatening the lives of passengers.

[0003] Wayfinding signage is a crucial facility for guiding people during evacuation. Currently, subway stations generally use static wayfinding signage, whose directions are fixed from the initial construction stage. However, in dynamically evolving fire scenarios, the fire, smoke, and crowd distribution all change in real time. Static wayfinding signage cannot respond in a timely manner and may continue to lead passengers to dangerous areas, thereby increasing evacuation risks. Therefore, dynamic wayfinding systems that can adjust their directions according to the real-time environment have become a potential means to improve evacuation efficiency.

[0004] However, dynamic guidance faces numerous technical challenges in practical applications. Fire environments are highly uncertain, with parameters such as temperature and visibility changing rapidly, making guidance decisions extremely complex. Furthermore, subway stations have numerous evacuation nodes, resulting in high-dimensional discrete characteristics in guidance decisions. Therefore, there is an urgent need to develop a dynamic and refined method for passenger evacuation in subway stations under uncertain fire environments. This method can collaboratively ensure evacuation efficiency and personnel safety under the most unfavorable disaster conditions, comprehensively improving the reliability and engineering applicability of subway station emergency evacuation systems. Summary of the Invention

[0005] To address the aforementioned technical problems, this disclosure provides a method for optimizing dynamic wayfinding signs.

[0006] In a first aspect, embodiments of this disclosure provide a dynamic wayfinding sign optimization method, including:

[0007] The internal evacuation space of the target subway station is divided into multiple cells, and the connectivity between multiple cells is established based on the actual spatial structure of the target subway station, the preset passage connection relationship, and the preset evacuation. The spatial discrete model of the target subway station is constructed, where the cells are used to represent local evacuation areas inside the subway station.

[0008] Define the number of passengers in each cell within the current time step and the passenger transmission flow between adjacent cells, and deduce the evacuation process of passengers in multiple cells based on preset supply and demand constraints to construct a cell transmission model.

[0009] Obtain fire environment parameters at different times within the target subway station, map the fire environment parameters to the cell level, and correct the passage capability of each cell in the cell transmission model under the fire environment.

[0010] A dynamic wayfinding sign optimization model is constructed with the objectives of cumulative passenger evacuation time, fire risk exposure level and density balance in key areas. In the dynamic wayfinding sign optimization model, the dynamic wayfinding sign decision variable represents the pointing state of each wayfinding sign in the current time step, which is used to control the flow path of passengers between cells.

[0011] The uncertainties in the fire environment are determined based on the fire environment parameters, and these uncertainties are embedded into the dynamic wayfinding sign optimization model to form a dynamic wayfinding sign robust optimization model.

[0012] The robust optimization model of dynamic wayfinding signage is solved, and the optimal dynamic wayfinding signage pointing strategy of the target subway station under the uncertain fire environment is obtained through iterative updates.

[0013] Secondly, embodiments of this disclosure provide a dynamic wayfinding sign optimization device, comprising:

[0014] The first model construction unit is used to divide the internal evacuation space of the target subway station into multiple cells, and establish the connectivity between multiple cells according to the actual spatial structure of the target subway station, the preset passage connection relationship and the preset evacuation, to construct the spatial discrete model of the target subway station. Here, the cells are used to represent the local evacuation areas inside the subway station.

[0015] The second model building unit is used to define the number of passengers in each cell within the current time step and the passenger transmission flow between adjacent cells, and to deduce the evacuation process of passengers in multiple cells based on preset supply and demand constraints, and to build a cell transmission model.

[0016] The model correction unit is used to obtain fire environment parameters at different times in the target subway station and map the fire environment parameters to the cell level to correct the passage capability of each cell in the cell transmission model in the fire environment.

[0017] The third model building unit is used to build a dynamic wayfinding sign optimization model with the objectives of cumulative passenger evacuation time, fire risk exposure level and density balance in key areas. The dynamic wayfinding sign decision variable in the dynamic wayfinding sign optimization model represents the pointing state of each wayfinding sign in the current time step, which is used to control the flow path of passengers between cells.

[0018] The robust optimization unit is used to determine the uncertainties in the fire environment based on the fire environment parameters, and to embed the uncertainties into the dynamic wayfinding sign optimization model to form the dynamic wayfinding sign robust optimization model.

[0019] The strategy update unit is used to solve the robust optimization model of dynamic wayfinding signs and obtain the optimal dynamic wayfinding sign pointing strategy of the target subway station under the uncertain fire environment through iterative updates.

[0020] Thirdly, embodiments of this disclosure provide an electronic device, including:

[0021] Memory;

[0022] Processor; and

[0023] Computer programs;

[0024] The computer program is stored in memory and configured to be executed by a processor to implement the first aspect of the method described above.

[0025] Fourthly, embodiments of this disclosure provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect above.

[0026] The dynamic wayfinding sign optimization method disclosed herein includes: dividing the internal evacuation space of a subway station into multiple cells, and constructing a spatial discrete model based on the actual spatial structure, preset channel connection relationships, and preset evacuation; defining the number of passengers in each cell and the passenger transmission flow between adjacent cells within the current time step, and deduce the evacuation process of passengers in multiple cells based on preset supply and demand constraints to construct a cell transmission model; mapping fire environment parameters to the cell level to correct the cell's passage capacity under fire conditions; constructing a dynamic wayfinding sign optimization model, and embedding the uncertainties under fire conditions into the dynamic wayfinding sign optimization model to form a robust optimization model for dynamic wayfinding signs; solving the robust optimization model, and obtaining the optimal dynamic wayfinding sign pointing strategy through iterative updates. The method provided in this application is applied to a subway station fire emergency evacuation system. By acquiring the evacuation operation status and fire environment information within the subway station in real time, it dynamically updates the wayfinding sign pointing scheme, guiding passengers to avoid high-risk areas and optimizing crowd distribution, thereby improving the overall evacuation efficiency and operational safety of the subway station under uncertain fire conditions. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0028] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 A flowchart illustrating a dynamic wayfinding sign optimization method provided in this embodiment of the disclosure;

[0030] Figure 2 A schematic diagram of a spatial discrete model provided in an embodiment of this disclosure;

[0031] Figure 3 A schematic diagram of a fire dynamics model provided in this embodiment of the present disclosure;

[0032] Figure 4 A schematic diagram illustrating the solution of a robust optimization model provided in an embodiment of this disclosure;

[0033] Figure 5 This is a schematic diagram of the structure of a dynamic wayfinding sign optimization device provided in an embodiment of the present disclosure;

[0034] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0035] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0036] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.

[0037] Specifically, with the accelerating pace of urbanization and the rapid development of urban rail transit systems, subways, with their advantages of large capacity, high efficiency, and punctuality, have gradually become one of the most important modes of transportation in urban rail transit systems, characterized by highly concentrated passenger flow. However, subway stations are generally characterized by enclosed spaces, complex structures, and dense crowds. In the event of a fire, smoke spreads rapidly, visibility decreases significantly, and localized congestion can easily occur, delaying passenger evacuation and posing a serious threat to human life. Especially in large or transfer subway stations, the complex spatial layout and limited evacuation routes further amplify the evacuation risks in fire scenarios. Therefore, developing efficient and reliable evacuation guidance methods in fire scenarios has become an important research topic for ensuring the safe operation of subway systems.

[0038] Wayfinding signage, as a crucial facility guiding passengers in route selection, directly impacts evacuation efficiency and risk control. Existing subway stations mostly use static wayfinding signage, whose direction is fixed during the design phase and difficult to adjust according to dynamic changes in fire development and crowd distribution. In complex fire environments, static wayfinding signage may continuously guide passengers into high-risk or highly congested areas, leading to route selection errors, further exacerbating local congestion and increasing risk exposure levels. In contrast, dynamic wayfinding signage, capable of adjusting its direction in real time based on environmental information and crowd conditions, is considered an effective means of improving the safety and efficiency of fire evacuation in subway stations.

[0039] However, the application of dynamic wayfinding signage in real-world fire scenarios still faces numerous challenges. Factors such as temperature, visibility, and toxic gas concentrations in a fire environment change rapidly over time and space, leading to significant uncertainty in wayfinding signage decisions. Passenger evacuation behavior exhibits obvious time-varying and spatially coupled characteristics, with wayfinding signage guidance and crowd flow mutually influencing each other, easily triggering amplified local congestion. Furthermore, subway stations have complex spatial structures and numerous evacuation nodes, resulting in high-dimensional discreteness in dynamic wayfinding signage decisions. Ensuring evacuation efficiency while avoiding frequent route changes and congestion accumulation becomes a key challenge in the design and optimization of dynamic wayfinding signage. Therefore, it is urgent to construct a dynamic and refined modeling method for passenger evacuation in subway stations under uncertain fire environments, and based on this, to achieve robust optimization decisions for wayfinding signage. This will enable coordinated protection of evacuation efficiency and personnel safety under the most unfavorable disaster conditions, comprehensively improving the reliability and engineering applicability of subway station emergency evacuation systems.

[0040] To address the aforementioned technical problems, this disclosure provides a dynamic wayfinding sign optimization method. First, based on the actual spatial layout of the subway station, the connectivity of passageways, and historical passenger flow distribution, a spatial discrete cellular model is constructed to form an evacuation network. Next, a cellular transmission model is used to establish a dynamic passenger evacuation simulation framework, defining cell capacity, supply-receiver capacity constraints, and a multi-downstream flow allocation mechanism based on evacuation priorities. Simultaneously, through fire dynamics simulation, fire environmental parameters such as temperature, visibility, and CO concentration are collected and mapped to construct a dynamic fire impact factor and a quantification mechanism for fire risk exposure levels. Based on this, with cumulative passenger evacuation time, fire risk exposure levels, and the balance of personnel density in key areas as multiple objectives, dynamic wayfinding signs are introduced as real-time controllable decision variables to construct a robust optimization model for dynamic wayfinding signs that includes box-type uncertainty sets and cross-budget uncertainty sets. To efficiently solve the robust optimization model based on real-time fire scenarios and crowd status information, a genetic algorithm improved from the kangaroo escape optimization algorithm is employed. By combining global search with Zigzag adaptive escape-style local perturbation, the directional signage pointing scheme is iteratively updated, ultimately obtaining the optimal dynamic directional signage pointing strategy that balances evacuation efficiency and safety even under the most unfavorable fire scenario. This application effectively improves the reliability and intelligence level of emergency evacuation guidance in subway stations under complex fire environments through the deep integration of high-resolution dynamic modeling, robust fire risk quantification, and intelligent optimization algorithms. Detailed explanations are provided through one or more of the following embodiments.

[0041] The dynamic wayfinding sign optimization method provided in this disclosure is applicable to dynamic wayfinding sign optimization scenarios. This method can be executed by a dynamic wayfinding sign optimization device, which can be implemented in software and / or hardware and can be integrated into an electronic device. The electronic device can include, but is not limited to, mobile terminals such as smartphones, laptops, digital radio receivers, personal digital assistants (PDAs), tablet PCs, PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), wearable devices, etc., as well as fixed terminals such as digital televisions, desktop computers, smart home devices, etc.

[0042] Figure 1 This is a flowchart illustrating a dynamic wayfinding sign optimization method provided in an embodiment of the present disclosure, specifically including as follows: Figure 1 The following steps are shown:

[0043] S101. Divide the internal evacuation space of the target subway station into multiple cells, and establish the connectivity between multiple cells based on the actual spatial structure of the target subway station, the preset channel connection relationship and the preset evacuation, and construct the spatial discrete model of the target subway station.

[0044] In this context, cells are used to represent local evacuation areas inside subway stations.

[0045] Understandably, the internal evacuation space of the target subway station is discretized and modeled to obtain a spatial discrete model. Specifically, the internal evacuation space of the subway station is divided into several square discrete cells according to a preset scale. Each cell represents a local evacuation area within the subway station. Connectivity relationships between cells are established based on evacuation direction constraints, the actual spatial structure, and the connection relationships of evacuation channels, forming a subway station evacuation network composed of cell nodes and their connectivity relationships. This completes the construction of the spatial discrete model, such as... Figure 2 As shown in the figure. Connectivity includes the adjacency relationships between cells and the traversable directions. Furthermore, different types of cells can be classified into ordinary evacuation cells, starting cells, and exit cells based on their functional attributes.

[0046] S102. Define the number of passengers in each cell within the current time step and the passenger transmission flow between adjacent cells, and deduce the evacuation process of passengers in multiple cells based on preset supply and demand constraints to construct a cell transmission model.

[0047] Understandably, based on the above S101, the passenger evacuation process in the subway station is dynamically modeled to obtain the Cellular Transport Model (CTM). The passenger transport flow (passenger quantity status) in the discrete time step and the passenger transport flow between cells are defined. Through the constraints of supply capacity and receiving capacity and the principle of flow conservation between cells, the spatiotemporal evolution process of passengers in the evacuation network is deduced, which is to say, the passenger evacuation process in the subway station is dynamically modeled.

[0048] Among them, multiple cells include the first cell and the second cell, and the first cell is the adjacent upstream cell of the second cell.

[0049] Understandably, multiple cells include the first cell and the second cell, where the first cell and the second cell are adjacent cells, and the first cell is the upstream cell of the second cell. The upstream cell refers to the cell in the first cell where passengers will be transferred to the second cell, that is, passengers will move from the current cell (the first cell) to the adjacent downstream cell (the second cell).

[0050] Optionally, the number of passengers in each cell within the current time step and the passenger transfer flow between adjacent cells are defined. Based on preset supply and demand constraints, the evacuation process of passengers in multiple cells is deduced, and a cell transfer model is constructed, including:

[0051] Based on the physical property parameters of the cells and the free movement speed of passengers in free movement state, calculate the current time step; define the number of passengers in each cell within the current time step; calculate the supply capacity of the first cell within the time step based on the number of passengers in the first cell, the cell area, and the first preset passage conditions, where the supply capacity is used to characterize the maximum number of passengers output by the first cell to the second cell in the current time step; calculate the receiving capacity of the second cell within the time step based on the remaining available space of the second cell, the cell area, and the second preset passage conditions, where the receiving capacity is used to characterize the maximum number of passengers received by the second cell from the first cell in the current time step; based on the free movement speed, preset cell critical density, preset cell congestion density, and preset congestion movement speed, deduce the evolution characteristics of the passenger flow transmission within the cells over time during passenger evacuation to obtain the actual cell density; define the passenger transmission flow between adjacent cells within the current time step based on the actual cell density, supply capacity, and receiving capacity, and construct the cell transmission model.

[0052] Understandably, the evacuation space of the subway station is divided into square discrete units. Each cell is used to represent a local evacuation area in the subway station. Each cell is assigned a maximum capacity, physical attribute parameters, and a time step (also called a discrete time step). The physical attribute parameters include the distance between the centers of adjacent cells. The calculation formula for the time step is shown in formula (1).

[0053] Formula (1)

[0054] In the formula, For time steps, The distance between the centers of adjacent cells. This indicates the speed of movement of a group of people in a state of free motion, i.e., the speed of free motion.

[0055] Understandably, at each discrete time step, a state variable representing the number of passengers in each cell is defined. This variable is used to characterize the passenger stock status in each cell at the current time step. At the same time, a flow rate representing the actual passenger flow between adjacent cells is defined. The flow rate between cells is constrained by both supply and receiving capacity. Under the premise of satisfying the flow conservation relationship, a dynamic description of the evolution of passengers in the subway station evacuation network over time is achieved, as shown in formula (2).

[0056] Formula (2)

[0057] In the formula, Represents cells within a time step (Referred to as the first cell) to the cell The actual number of people transmitted (denoted as the second cell), This indicates the cell's supply capacity within a time step. This indicates the cell's supply capacity within a time step.

[0058] Understandably, the calculation process for the supply capacity of the first cell includes: calculating the supply capacity of the upstream cell at the current time step based on the current number of passengers, cell area, and passage conditions, as shown in formula (3). Here, the supply capacity is used to characterize the maximum number of passengers that the cell can output to its adjacent downstream cells at the current time, under the premise of satisfying the cell maximum capacity constraint and passage capacity limit.

[0059] Formula (3)

[0060] In the formula, Represents cell Supply capacity Indicates time step cell The density (i.e., the number of passengers in the first cell). Represents cell area, This represents the flow-density relationship.

[0061] Understandably, the calculation process for the receiving capacity of the second cell includes: calculating the receiving capacity of the downstream cell at the current time step based on the remaining available space and passage conditions of the downstream cell, as shown in formula (4). Here, the receiving capacity is used to characterize the maximum number of passengers that the downstream cell can receive from the adjacent upstream cell at the current time, provided that the maximum capacity constraint of the downstream cell is not exceeded.

[0062] Formula (4)

[0063] In the formula, Represents a cell The receiving capability, Represents a time step cell The number of people who left.

[0064] Understandably, the supply and receiving capacity of each cell at the current time step are used as constraints to determine the actual passenger flow between adjacent cells, that is, the actual number of passengers in the cell after the passenger flow is completed within the time step. Subsequently, the passenger flow is allocated under the premise of not exceeding the corresponding supply and receiving capacity, and the passenger number state variable of each cell is updated to ensure that the passenger flow between cells satisfies the number conservation relationship, as shown in formula (5).

[0065] Formula (5)

[0066] In the formula, Indicates time step Inner cells The number of people who left Indicates time step Inner cells density, Indicates time step influx into cells The number of people.

[0067] Understandably, a specific flow-density relationship is used in the Cellular Transport Model (CTM) to calculate the passenger flow between cells, as shown in Equation (6). This relationship uses a piecewise function to characterize the nonlinear dynamic characteristics of crowd flow in a fire environment, ensuring that CTM can accurately characterize the evolution of crowd flow over time during evacuation and obtain the actual density of the cells.

[0068] Formula (6)

[0069] In the formula, It is the free flow velocity. It is the critical density of the cell. It is the cell congestion density. It refers to the speed at which passengers move in congested traffic.

[0070] Understandably, based on the actual density of the cells, the passenger transmission flow between adjacent cells within the current time step is defined to satisfy the constraints of supply capacity and receiving capacity, thus obtaining the cell transmission model.

[0071] Optionally, define the passenger transport flow between adjacent cells within the current time step, including:

[0072] When the first cell is connected to multiple second cells in the current time step, the evacuation field value corresponding to each second cell is calculated, where the evacuation field value is used to characterize the evacuation superiority in the downstream direction; the allocation ratio of the crowd transmission flow to each second cell is determined according to the evacuation field value corresponding to each second cell; based on the allocation ratio of each second cell, the passenger transmission flow between adjacent cells in the current time step is defined according to the evacuation priority of each second cell.

[0073] Understandably, in passenger evacuation within a subway station, when an upstream cell corresponds to multiple passable downstream cells in the current time step, a multi-downstream crowd flow allocation mechanism based on evacuation priority is introduced, that is, passengers are preferentially allocated to downstream cells with better evacuation conditions. The following embodiment uses the first cell corresponding to three second cells as an example for illustration. First, the evacuation field value of the three downstream cells in the current time step is calculated, as shown in formula (7). This evacuation field value reflects the superiority of evacuation in the downstream direction, so as to weightedly allocate the crowd transmission flow according to the evacuation field value corresponding to each downstream cell.

[0074] Formula (7)

[0075] In the formula, For the current time step The evacuation field value of the downstream cell.

[0076] Understandably, the flow of people first flows to the downstream cell with the lowest evacuation field value. When its receiving capacity is insufficient to absorb the remaining people in the upstream cell, the remaining people flow to the second-best direction cell and the third-best direction cell in sequence, as shown in formula (8). This mechanism ensures the physical consistency and rationality of the flow, and can effectively avoid congestion and convection caused by sudden changes in flow direction or excessive concentration.

[0077] Formula (8)

[0078] In the formula, , and Indicates time step upstream cells Downstream cells , and The actual number of people transmitting, Indicates time step upstream cells Supply capacity , and Indicates time step Downstream cells , and The demand for capability.

[0079] Understandably, at the end of each time step, the passenger quantity state variable of each cell is updated based on the actual passenger flow between adjacent cells during that time step. Combined with the cell's passage capacity and the dynamic impact of fire environmental risk on the evacuation process, the passenger stock status of each cell in the next time step is obtained, thus forming a dynamic evacuation process that reflects the evolution of passenger evacuation status in the subway station over time.

[0080] S103. Obtain fire environment parameters at different times within the target subway station, map the fire environment parameters to the cell level, and correct the passage capability of each cell in the cell transmission model under fire conditions.

[0081] Understandably, based on the above S102, after a fire occurs, its development process exhibits significant temporal evolution characteristics. Key factors such as fire intensity, smoke diffusion, and ambient temperature change continuously over time, exerting a sustained impact on the feasibility of passenger evacuation. In this context, fire environmental parameters such as temperature, visibility, and CO concentration at different times within the subway station are obtained through fire dynamics model simulation. A fire risk exposure level quantification mechanism is then constructed for the subway station fire evacuation scenario. This mechanism maps fire environmental parameters to the cellular level of the cellular transport model, correcting the passage capacity of each cell in the model to characterize the impact of fire environmental evolution on cell passage capacity and passenger evacuation behavior. In other words, a dynamic correction mechanism for the fire environment's impact on cell passage capacity is introduced into the cellular transport model. The fire risk exposure level quantification mechanism normalizes each fire environmental parameter and assigns corresponding risk weight coefficients, comprehensively representing the impact of different environmental factors on passenger evacuation safety. The fire dynamics model is as follows: Figure 3 As shown.

[0082] Optionally, fire environment parameters at different times within the target subway station are obtained, and these parameters are mapped to the cell level to correct the mobility of each cell in the cell transport model under fire conditions, including:

[0083] The fire environment parameters at different times within the target subway station are obtained and mapped to the cell level of the cell transport model to obtain the fire environment state corresponding to each cell at different time steps. Based on the fire environment state, a fire environment influence factor is constructed, which is used to quantify the inhibitory effect of the fire environment on passenger passage capacity. The passage capacity of each cell in the free flow state is corrected according to the fire environment influence factor to obtain the corrected cell transport model, which takes into account the effective passage capacity of the cell after the fire environment influence.

[0084] Understandably, fire environmental parameters of different areas within a subway station at different times can be obtained through fire dynamics simulation or environmental monitoring. These parameters include temperature, visibility, and CO concentration. Specifically, simulation software is used to obtain the temporal variation patterns of key pyrophysiological parameters during fire development and to solve for gas flow and energy transfer during the fire. Under the Fire Dynamics Simulator numerical simulation framework, fire development can be reduced to the numerical solution of the gas flow mass conservation, energy conservation, and momentum conservation equations. The governing equations are the continuity equation, energy equation, and momentum equation, as shown in equations (9) to (11).

[0085] Formula (9)

[0086] Formula (10)

[0087] Formula (11)

[0088] In the formula, , and Representing gas density, velocity vector, and visible enthalpy respectively; pressure From the ideal gas law Sure; and Representing the heat release rate and heat flux vector per unit volume, respectively. It is a dissipation term; , and These represent the gravity term, the external force term, and the Newtonian fluid viscous stress tensor, respectively.

[0089] Furthermore, adopting the classic A fire model, as the input condition for the fire source, is used to describe the intensity growth process of a fire from its initial stage to its development stage. To balance the rationality of the fire evolution description with the feasibility of path optimization calculations, the following... The fire is set as a simplified fire intensity model, as shown in formula (12).

[0090] Formula (12)

[0091] in, Indicates time heat release rate, This is the fire growth coefficient, used to reflect the rate of fire development. This indicates the time when the fire intensity reaches its peak.

[0092] Understandably, the influence of fire changes over time can also be incorporated into the robust optimization model for dynamic directional signage based on the fire intensity evolution model (fire dynamics model), thereby achieving an effective connection between fire dynamics simulation results and evacuation decision models.

[0093] Understandably, after determining the fire environment parameters through simulation, the fire environment parameters are mapped to the cell level of the cell transport model to obtain the fire environment state corresponding to each cell at different time steps. Subsequently, based on this, a fire environment influence factor is constructed based on the fire environment state, whereby the fire environment influence factor is used to quantify the inhibitory effect of the fire environment on passenger passage capacity. And the passage capacity of each cell in the free flow state is corrected according to the fire environment influence factor to obtain the corrected cell transport model, so that the cell transport model can reflect the actual characteristics of the evolution of crowd evacuation capacity with time and spatial distribution under fire conditions, as shown in formula (13).

[0094] Formula (13)

[0095] In the formula, Indicates time step in fire situation Density-flow ratio, This represents the flow rate correction factor in a fire scenario. Indicates time step Density-flow ratio, Indicates time step Inner cells The density.

[0096] Understandably, the effective throughput capacity of cells is adjusted based on fire environmental impact factors, and this adjusted capacity is embedded into the supply capacity and flow calculation process of the cell transmission model, allowing the flow of people between cells to be dynamically adjusted as the fire environment changes. Subsequently, through a dynamic adjustment mechanism, the time-varying evolution of cell throughput capacity under fire conditions is realized, thereby reflecting the continuous impact of the fire environment on the passenger evacuation process.

[0097] Optionally, fire environmental impact factors can be constructed based on the fire environmental conditions, including:

[0098] The quantized value of visibility at a cell is calculated based on a preset weighting coefficient for visibility, the actual value of visibility at the cell within the current time step, and the initial value of visibility at the cell within the current time step. A visibility factor is then calculated based on this quantized value, whereby the visibility factor characterizes the impairment of pedestrian mobility caused by smoke-induced visual impairment. Similarly, the quantized value of CO concentration at a cell is calculated based on a preset weighting coefficient for CO concentration, the actual value of CO concentration at the cell within the current time step, and the initial value of CO concentration at the cell within the current time step. A visibility factor is then calculated based on this quantized value. The CO concentration factor is calculated, which characterizes the impact of the cumulative effect of CO exposure on the risk of poisoning on the mobility of the population. The quantified value of the temperature at the cell is calculated based on the preset weighting coefficient of temperature, the actual value of the temperature at the cell in the current time step, and the initial value of the temperature at the cell in the current time step. The temperature factor is calculated based on the quantified value of the temperature at the cell, which characterizes the constraint of thermal radiation damage and heat resistance limit caused by high temperature environment on the mobility of the population. The fire environment impact factor is constructed based on the visibility factor, CO concentration factor, and temperature factor.

[0099] Understandably, in a fire scenario, factors such as smoke, CO concentration, and high temperature can reduce the visibility and mobility of pedestrians, thereby weakening the passage capacity of cells. Therefore, the flow rate correction coefficient between cells in a fire scenario is embedded in the CTM and used as a multiplier of the original flow rate of the CTM to obtain the actual flow rate affected by the fire. The calculation method of the above flow rate correction coefficient is shown in formula (14).

[0100] Formula (14)

[0101] In the formula, This is the flow rate correction factor. Visibility factor CO concentration factor The temperature factor is used to describe the impairment of mobility caused by visual impairment due to smoke, as shown in formula (15). The CO concentration factor is used to characterize the impact of the cumulative effect of CO exposure on mobility due to poisoning risk, as shown in formula (16). The temperature factor is used to reflect the constraint of high temperature environment on mobility due to heat radiation damage and heat resistance limit, as shown in formula (17).

[0102] Formula (15)

[0103] Formula (16)

[0104] Formula (17)

[0105] In the formula, Representing time steps Zhongyuanbao The quantitative values ​​of temperature, visibility and CO concentration are calculated using formulas (18) to (20).

[0106] Understandably, by quantifying the cumulative effect of fire risk experienced by passengers in each cell during evacuation, the safety guidance capability of dynamic signage and guidance strategies in complex fire environments can be characterized. Specifically, firstly, the degree of danger of the fire environment is quantitatively modeled, as shown in formulas (18) to (20).

[0107] Formula (18)

[0108] Formula (19)

[0109] Formula (20)

[0110] In the formula, Representing time steps Zhongyuanbao Quantitative values ​​of temperature, visibility, and CO concentration. These represent the weighting coefficients for temperature, visibility, and CO concentration, respectively, and are set to 0.3, 0.5, and 0.2. Representing time steps Zhongyuanbao The values ​​of temperature, visibility, and CO concentration were recorded. Representing time steps Zhongyuanbao The initial values ​​for temperature, visibility, and CO concentration were 20 °C, 10 m, and 0.0005 mol / mol, respectively.

[0111] S104. Construct a dynamic wayfinding sign optimization model with the objectives of cumulative passenger evacuation time, fire risk exposure level, and density balance in key areas.

[0112] In the dynamic wayfinding sign optimization model, the dynamic wayfinding sign decision variable represents the directional state of each wayfinding sign in the current time step, which is used to control the flow path of passengers between cells.

[0113] Understandably, based on the above S103 and the modified cellular transport model, a dynamic guidance sign decision variable is introduced. The cumulative evacuation time of passengers, the level of fire risk exposure, and the balance of personnel density in key areas are used as optimization objectives to comprehensively evaluate evacuation efficiency, safety risks, and population distribution. A dynamic guidance sign optimization model is constructed, in which the dynamic guidance sign variable is used to represent the pointing state of the guidance sign in the discrete time step, and the dynamic guidance of passenger evacuation paths is realized by controlling the passable direction between cells.

[0114] Understandably, several dynamic directional signs are installed in subway stations, each capable of switching between different directions in real time to guide passengers to choose the optimal evacuation route. The directional state of these signs is determined by a binary variable. It means that, among them, Represents the first dynamic wayfinding sign. This indicates its optional pointing direction. When When, it indicates that the identifier currently points to a cell. The direction is the passable path; when When the time is specified, it indicates that the direction is closed or not selected. The overall pointing configuration of all dynamic wayfinding signs forms a state set. Different combinations of signs correspond to different directional strategies, enabling real-time control of crowd flow paths. Furthermore, while the dynamic directional signage system has real-time switching capabilities, to avoid passenger confusion or decision-making disarray due to frequent changes in instructions, its directional state is only updated uniformly at the beginning of each time step and remains fixed within that time step. Subsequent passengers strictly follow the guidance of the dynamic directional signs, choosing the passable direction indicated by the sign as their evacuation route, exhibiting no obvious reverse movement, thus ensuring the orderliness and predictability of crowd flow in the model.

[0115] Optionally, a dynamic wayfinding sign optimization model can be constructed with the optimization objectives of cumulative passenger evacuation time, fire risk exposure level, and balanced population density in key areas.

[0116] Understandably, to improve the safety and efficiency of passenger evacuation in subway stations during fires, a multi-objective evacuation performance evaluation function, also known as a dynamic guidance sign optimization model, is constructed based on the cellular transport model. Specifically, using the CTM as the basic framework for crowd movement, robust optimization and control of guidance signs are carried out from three dimensions: evacuation efficiency, safety risk, and load balancing of key nodes. A multi-objective evacuation performance evaluation function is constructed, as shown in formula (22), where the objective functions correspond to the cumulative passenger dwell time, fire risk exposure level, and the degree of balance of personnel density in key areas, respectively. Based on the multi-objective evacuation performance evaluation function, the uncertainty of fire environment risk parameters is introduced to constrain and optimize the evacuation performance of guidance sign pointing strategy under the most unfavorable disturbance conditions, thereby obtaining the optimal dynamic guidance sign pointing strategy that balances evacuation efficiency and passenger safety in an uncertain fire environment.

[0117] Optionally, a dynamic wayfinding signage optimization model is constructed, targeting cumulative passenger evacuation time, fire risk exposure level, and density balance in key areas, including:

[0118] The cumulative passenger dwell time is obtained by summing the passenger dwell times in each cell during the evacuation process. The quantified values ​​of visibility, CO concentration, and temperature at each cell are normalized and weighted to obtain the nominal risk value of each cell in the current time step. The nominal risk value is used to characterize the level of environmental risk experienced by passengers during the evacuation process. Based on the actual cell density in the key areas of the target subway station in the current time step and the average density of all key areas in the current time step, a passenger density variance function for the key areas is constructed. Based on the cumulative passenger dwell time, nominal risk value, and passenger density variance function, a dynamic wayfinding signage optimization model is constructed.

[0119] Understandably, the process of establishing the multi-objective evacuation performance evaluation function is as follows. The fire risks experienced by passengers in each cell during the evacuation process are cumulatively calculated, and the safety guidance performance of the directional signage strategy in the fire environment is evaluated. Specifically, the degree of danger of the fire environment is quantitatively modeled by superimposing the fire risk field. The risk fields (quantified values) corresponding to the temperature field, visibility field and CO concentration field are normalized and superimposed according to weights to obtain the risk value at each cell in the current time step, as shown in formula (21). This risk value is used to characterize the level of environmental risk experienced by passengers during the evacuation process. In addition, this risk value changes dynamically with time and space to reflect the time-varying characteristics of risk distribution during the fire evacuation process and serves as the risk input for dynamic directional signage optimization and robust modeling.

[0120] Formula (21)

[0121] In the formula, For time step cell The risk value (nominal risk value) at the location.

[0122] Understandably, based on the deviation between the nominal risk value and the risk value at the current time step, the maximum risk value allowed for the deviation at the current time step is obtained, as shown in formula (22).

[0123] Formula (22)

[0124] In the formula, For the first Each cell at time step The maximum allowable risk value for risk deviation. For the first Each cell at time step The nominal risk value, For the first Each cell at time step The deviation of the risk value.

[0125] Understandably, the actual cell density of cells in the key areas within the target subway station is determined at the current time step. The average density of cells in all key areas at the current time step is calculated, as shown in formula (23). Subsequently, based on the actual cell density and the average density, the passenger density variance function of the key areas at the current time step is constructed, where the passenger density variance function is used to reflect the degree of passenger evacuation.

[0126] Formula (23)

[0127] In the formula, For the key evacuation area of ​​the subway station, cells are gathered, and , This is a key evacuation area for subway stations. For key areas Inner Each cell at time step density, For key areas Inner cells at time step The average density.

[0128] Understandably, the dynamic guidance signage scheme is optimized by comprehensively evaluating the operational efficiency, safety risks and crowd distribution during the passenger evacuation process, as shown in formula (24).

[0129] Formula (24)

[0130] In the formula, , , The weights of the factors, and . The total number of cells inside the subway station. The maximum number of discrete time steps considered in the simulation process. For the first Each cell at time step The number of people.

[0131] S105. Determine the uncertainties in the fire environment based on the fire environment parameters, and embed the uncertainties into the dynamic wayfinding sign optimization model to form a dynamic wayfinding sign robust optimization model.

[0132] Understandably, based on the above S104, in view of the uncertainty of fire environmental risk parameters in time and space, a robust optimization method is introduced to model the cell risk value as an uncertain parameter, and the risk fluctuation range is constrained by the box uncertainty set and the cross budget uncertainty set, so as to embed uncertainty into the dynamic guidance sign optimization model, thereby forming a dynamic guidance sign robust optimization model that still has stability under the most unfavorable fire disturbance conditions.

[0133] Optionally, uncertainties in the fire environment are determined based on fire environment parameters, and these uncertainties are embedded into the dynamic wayfinding sign optimization model to form a robust optimization model for dynamic wayfinding signs, including:

[0134] Fire environmental parameters are set as uncertain parameters, which characterize the impact of fire development and environmental disturbances on the evacuation operation of the subway station. An uncertainty set is constructed for the uncertain parameters, which describes the fluctuation range of the uncertain parameters within the allowable range. The uncertainty set includes at least a box-type uncertainty set and a cross-calculation uncertainty set. Under the constraints of the uncertainty set, the dynamic wayfinding sign optimization model is robustly processed to form a robust optimization model for dynamic wayfinding signs, which still satisfies the evacuation operation constraints under the most unfavorable conditions. The robust optimization model for dynamic wayfinding signs is transformed to convert the most unfavorable conditions contained therein into a set of deterministic constraints of equal order.

[0135] Understandably, the robust optimization process of dynamic wayfinding signs in subway stations is as follows: the fire environmental risk parameter is set as an uncertain parameter, which is used to characterize the impact of fire development and environmental disturbance on the evacuation operation status of the subway station. Subsequently, an uncertain set is constructed for the uncertain parameter, which is used to describe the fluctuation range of the uncertain parameter within the allowable range. The uncertain set includes at least a box-type uncertain set and a cross-calculation uncertain set, as shown in formula (25). Under the constraint of the uncertain set, the robust optimization model of dynamic wayfinding signs is robustened so that the optimization model still meets the evacuation operation constraints under the most unfavorable uncertain situation. By performing an equivalent transformation on the robust optimization model, the robust optimization problem containing the worst-case constraint is transformed into a deterministic optimization problem for subsequent solution of the optimal direction scheme of dynamic wayfinding signs, so that the optimization model can be solved by deterministic optimization methods while maintaining the original robustness requirements.

[0136] Formula (25)

[0137] In the formula, This is the upper bound of the box-type indeterminate set. These are the control parameters for the budget uncertainty set.

[0138] Optionally, the robust optimization model for dynamic guidance signage is transformed to convert the worst-case conditions it contains into a set of deterministic constraints of equal order, including:

[0139] Construct an inner-layer maximization problem function based on the population distribution state, uncertain parameters, and worst-case deviation value of the cells; construct a Lagrangian function based on non-negative Lagrange multipliers and the maximization problem function; solve a one-dimensional maximization problem of the Lagrangian function with respect to the worst-case deviation value to obtain the optimal value of the worst-case deviation; substitute the optimal value into the maximization problem function, and obtain the total evacuation risk increment under the worst-case scenario based on the optimal deviation contribution of all cells and the Lagrange multiplier terms; substitute the total evacuation risk increment into the pre-constructed minimization problem function to obtain a dynamic guided sign robust optimization model containing a set of deterministic constraints.

[0140] Understandably, to solve the robust optimization model of dynamic wayfinding signage, the uncertainty of the most unfavorable risk value deviation involved must be transformed into a solvable deterministic form. Given the wayfinding signage decisions and the crowd distribution, the number of passengers in each cell at different times is considered a known constant. Let the number of passengers in the first cell be a known constant. In the nth discrete time step The risk value of each cell is composed of its nominal risk value and uncertainty deviation. Under the given guidance decision, the most adverse effect of risk deviation on the system objective function can be expressed as an inner-level maximization problem of the risk deviation variable, and the maximization problem function is constructed as shown in formula (26).

[0141] Formula (26)

[0142] In the formula, To maximize the problem function, For passenger numbers, The deviation of the most unfavorable risk value represents the control parameters in the box uncertainty set. and budgetary uncertain set control parameters Within the permissible range, the risk deviation that results in the worst passenger evacuation performance, among which, . worst deviation Constrained in an uncertain set Within this set, the box-type uncertainty set intersects with the budget uncertainty set to characterize the possible range of risk fluctuations. To minimize the evacuation penalty, only the adverse direction that increases risk needs to be considered; therefore, under the budget uncertainty set, the deviation only takes the upper bound of the box-type uncertainty set, i.e. By considering the most unfavorable deviation in the calculation. Optimized wayfinding signage can achieve robust adaptation to fire risk disturbances.

[0143] Understandably, this is because the inner maximization problem is a linear programming problem with simple boundary conditions and a single coupled budget constraint, and the risk uncertainty set... As a convex set, according to strong duality theory, the inner problem can be equivalently transformed through its dual problem. To obtain its closed-form solution, the Lagrange relaxation method is adopted, and the budget constraint is introduced into the Lagrange function for processing, transforming the optimization problem with coupled constraints into a separable unconstrained problem. Specifically, the budget constraint is dualized by introducing non-negative Lagrange multipliers, and then the Lagrange function is constructed, as shown in formula (27).

[0144] Formula (27)

[0145] In the formula, It is a non-negative Lagrange multiplier.

[0146] Understandable, for a given Value, the Lagrange function with respect to the deviation vector The maximization problem can be decomposed into maximizing each component. This is an independent one-dimensional maximization problem. For a fixed time step... Fixed cell We need to solve formula (28) to obtain information about It is a one-dimensional linear function, as shown in formula (29).

[0147] Formula (28)

[0148] Formula (29)

[0149] In the formula, In other words, given a Lagrange multiplier And wayfinding signage decisions and crowd distribution Given a constant value, solve the Lagrange function with respect to the risk deviation value. After solving the one-dimensional maximization problem, the resulting risk bias is... The optimal value.

[0150] Understandably, by substituting the optimal value back into the inner maximization problem function, summing the optimal deviation contributions of all cells, and adding the Lagrange multipliers corresponding to the budget constraint, we obtain the final result of the inner maximization problem, which is the total evacuation risk increment in the worst case, as shown in formula (30).

[0151] Formula (30)

[0152] Formula (31)

[0153] In the formula, In other words, given a Lagrange multiplier Next, time step Cells that exhibit risk value deviations in the most unfavorable risk scenario A set of.

[0154] Understandably, after obtaining the closed form of the most unfavorable deviation of each cell, the inner maximization result can be substituted into the outer minimization problem function, thereby transforming the original max-min double nested structure into an equivalent single-layer optimization problem, as shown in formula (32).

[0155] Formula (32)

[0156] S106. Solve the robust optimization model of dynamic wayfinding signs and obtain the optimal dynamic wayfinding sign pointing strategy of the target subway station under the uncertain fire environment through iterative updates.

[0157] Understandably, based on the above S105, in the process of solving the robust optimization model, a genetic algorithm based on the improved kangaroo escape optimization algorithm is used to solve the robust optimization model. By combining global search and escape-type local perturbation, the discrete guidance sign pointing strategy is iteratively updated to obtain the optimal dynamic guidance sign pointing strategy that balances evacuation efficiency and passenger safety in the uncertain environment of a fire.

[0158] Optionally, a robust optimization model for dynamic wayfinding signage is solved, and the optimal dynamic wayfinding signage direction strategy for the target subway station under uncertain fire conditions is obtained through iterative updates, including:

[0159] The pointing strategy of the dynamic wayfinding sign given by the robust optimization model of the dynamic wayfinding sign is encoded as a discrete decision sequence in the current time step, which serves as the individual representation of each individual in the genetic algorithm. The parameters of the genetic algorithm are initialized and an initial population is generated. The fitness value of the wayfinding sign corresponding to each individual in the initial population is calculated according to the deterministic optimization model. Based on the fitness value, selection, crossover and mutation operations are performed on the individuals in the population to generate candidate dynamic wayfinding sign pointing strategies. During the iterative search in the discrete solution space, the candidate dynamic wayfinding sign pointing strategies are updated based on the kangaroo escape optimization algorithm to obtain the optimal dynamic wayfinding sign pointing strategy of the target subway station under the uncertain fire environment.

[0160] Understandably, for the robust optimization model of dynamic wayfinding signs, the direction of each wayfinding sign within a discrete time step is used as the decision variable, making the optimization model a high-dimensional discrete combinatorial optimization problem. Simultaneously, the model introduces an uncertainty set to describe the fluctuating characteristics of the fire environment and path risks, and uses worst-case constraints to characterize the evacuation safety performance requirements under adverse conditions, resulting in a complex solution space and nested robust constraints. To address these characteristics, a genetic algorithm improved from the kangaroo escape optimization algorithm is used as the global search framework to generate discrete decision sequences for wayfinding sign directions, and the robust subproblem of the uncertainty set under a given decision sequence is solved precisely. Specifically, the pointing strategy of the dynamic guide marker within a discrete time step is encoded as a discrete decision sequence, serving as the individual representation of the genetic algorithm. The genetic algorithm parameters are initialized and an initial population is generated. The fitness value is calculated for the guiding marker pointing scheme corresponding to each individual, provided by a deterministic optimization model. Based on the fitness value, selection, crossover, and mutation operations are performed on the individuals in the population to generate new candidate dynamic guide marker pointing strategies. During the iterative search process, a guidance mechanism from the kangaroo escape optimization algorithm is introduced to update the candidate solutions, enhancing the local search capability of the genetic algorithm in the discrete solution space. The above steps are repeated until a preset iteration termination condition is met, outputting the dynamic guide marker pointing strategy with the optimal fitness.

[0161] Understandably, the pointing scheme of dynamic guidance markers within a discrete time step is encoded as a discrete decision sequence, serving as the individual identifier for the genetic algorithm. Specifically, suppose that the guidance markers are set up in a subway station... Each of the three dynamic directional sign nodes is constrained by its spatial location, connectable channels, and evacuation path structure. It is only allowed to switch directions within a pre-defined, finite set of candidate directions, and selects the corresponding direction from the candidate direction set within a discrete time step. Let the first node be... The set of possible pointing directions for each directional sign is shown in formula (33), denoted as the first... The number of selectable directions for each directional sign is Then its corresponding decision variable at each discrete time step The value constraint on can be expressed as formula (34).

[0162] Formula (33)

[0163] Formula (34)

[0164] In the formula, Represents the set of decision variables. Indicates the first The first dynamic wayfinding sign in the Decision variables at each time step Indicates the first A set of selectable directions for dynamic wayfinding signs.

[0165] Understandably, at the encoding level, integer encoding is used to map the discrete decision vectors mentioned above to the chromosome structure in the genetic algorithm, where the chromosome length and set... The encoding set contains the feasible pointing direction numbers of the corresponding identifier nodes at each time step, with each gene position value having the same length. As shown in formula (35).

[0166] Formula (35)

[0167] Understandably, after completing the initial population construction, the genetic algorithm iteratively updates the guidance marker pointing schemes through selection, crossover, and mutation operations to generate candidate solution combinations with better evacuation performance within the discrete feasible solution space. Specifically, in the selection phase, based on the comprehensive fitness values ​​of each individual in the current population under indicators such as evacuation time, risk exposure, and crowding level, a roulette wheel selection strategy is used to construct a mating pool, allowing guidance marker pointing schemes with lower fitness to participate in subsequent genetic operations with a higher probability. In the crossover phase, pairs of individuals are randomly selected from the mating pool, and a crossover operation is performed on their guidance marker pointing sequences at the chromosome level. Let the parent chromosomes be... and Then the crossover operation can be expressed as formula (36).

[0168] Formula (36)

[0169] In the formula, This represents the discrete crossover operator performed under a given crossover probability.

[0170] Understandably, during the mutation phase, to enhance population diversity and avoid getting trapped in local optima during the search process, random perturbations are applied to the gene loci of some chromosome individuals. The first... A directional sign in time step The direction of decision variables Replace with its set of optional directions Another feasible value in the equation can be expressed by the variation rule as formula (37).

[0171] Formula (37)

[0172] In the formula, Indicates the first A directional sign in time step The current pointing direction, Indicates the first A directional sign in time step The candidate pointing direction.

[0173] Optionally, during the iterative search in the discrete solution space, the candidate dynamic wayfinding sign pointing strategies are updated based on the kangaroo escape optimization algorithm to obtain the optimal dynamic wayfinding sign pointing strategy for the target subway station under the uncertain fire environment, including:

[0174] A subset of individuals is selected from the population using a genetic algorithm, and the guidance markers corresponding to these individuals are used as initial solutions input into the kangaroo escape optimization algorithm. A direction vector pointing to the current optimal solution is constructed, ensuring that individuals have the ability to move to a better region. A random rotation is applied to the direction vector, and the rotation angle is calculated. Based on the unit vector and rotation angle, the rotated escape direction is generated. Based on the escape direction and preset jump strength parameters, a Gaussian perturbation term is used to implement nonlinear offset during the individual search process. A directional guidance term between the optimal individual and the current individual is introduced to encourage individuals to jump towards the current optimal solution. After continuous perturbation updates, the updated individuals are mapped back to a preset discrete domain using nearest neighbor projection to ensure that the updated perturbation maintains the escape search characteristics while satisfying the feasibility constraints of the dynamic guidance marker pointing strategy. The evacuation efficiency and safety performance of candidate dynamic guidance marker pointing strategies are comprehensively evaluated, and the population is selected and updated based on the comprehensive evaluation results to continue the next iteration until the termination condition is met, at which point the iteration stops, yielding the optimal dynamic guidance marker pointing strategy.

[0175] Understandably, for each set of dynamic guidance marker pointing decision sequences generated by the genetic algorithm in the current iteration, it is taken as the given solution of the outer optimization problem and substituted into the robust optimization model with uncertainty set constraints to form the corresponding inner risk assessment subproblem. Specifically, the Gurobi solver is used to solve the above robust subproblem precisely. After completing the global search based on the genetic algorithm and the robust risk assessment of the corresponding guidance marker pointing scheme, some individuals with high fitness are selected from the genetic algorithm population, and their corresponding guidance marker pointing decision sequences are used as the initial solution input into the kangaroo escape optimization algorithm module, and its Zigzag escape mechanism is used to perform fine perturbation update on the solution. The Zigzag strategy first constructs a direction vector pointing to the current optimal solution to ensure that the individual has the ability to move to a better region, as shown in formula (38). Then, in order to simulate the rapid change of direction of the kangaroo, the kangaroo escape optimization algorithm applies a random rotation to this direction, and its angle is calculated as shown in formula (39). Then, combined with the unit vector perpendicular to the direction vector, The generated escape direction after rotation is shown in Equation (40). Finally, the Zigzag update achieves nonlinear local search through the superposition of directional jumps and Gaussian perturbations, and its update rule is shown in Equation (41).

[0176] Formula (38)

[0177] Formula (39)

[0178] Formula (40)

[0179] Formula (41)

[0180] Understandably, in the hybrid optimization strategy, a perturbation update mechanism based on the kangaroo escape optimization algorithm is introduced, where a Gaussian perturbation term is set. To achieve nonlinear offset in the individual search process, a directional guiding term between the optimal individual and the current individual is introduced. This prompts individuals to jump towards the current optimal solution. This can be achieved by setting a jump strength parameter. By adjusting the perturbation amplitude and energy level during the individual search process, the individual can converge toward the optimal solution while also taking into account local search capabilities, thereby increasing the probability of escaping local optima.

[0181] Furthermore, to address the issues of slow convergence and insufficient local exploitation capabilities in genetic algorithms when approaching the optimal solution region, the perturbation update mechanism of the kangaroo escape optimization algorithm is embedded after the mutation step of the genetic algorithm. After completing the crossover and mutation operations, a perturbation update based on the kangaroo escape mechanism is performed on the offspring individuals, as shown in formula (42). This achieves a collaborative optimization process combining global search and local enhancement, thereby accelerating the search speed in the direction of the optimal solution and improving the overall optimization performance of the algorithm.

[0182] Formula (42)

[0183] Understandably, since the decision variables in the optimization model take values ​​of discrete finite sets, after performing continuous perturbation updates on individuals, the updated individuals are mapped back to the preset discrete domain through the nearest neighbor projection method, as shown in formula (43), so as to ensure that the solution after perturbation update maintains the escape search characteristics while satisfying the feasibility constraint of the dynamic guidance sign pointing decision.

[0184] Formula (43)

[0185] Understandably, after completing genetic operations, robust risk assessment, and local enhancement search, the evacuation efficiency and safety performance of each dynamic guidance marker pointing scheme are comprehensively evaluated. Based on the fitness results, the population is selected and updated, retaining the schemes with better performance for the next generation of evolution. During the iteration process, the algorithm convergence is determined based on the maximum number of iterations or the fitness change threshold. When the termination condition is met, the iteration stops, and the optimal dynamic guidance marker pointing scheme that balances evacuation efficiency and robust safety is output.

[0186] The dynamic wayfinding sign optimization method provided in this disclosure achieves high spatiotemporal resolution dynamic simulation of passenger evacuation processes under complex station structures by discretizing the subway station space into cells and combining it with a cellular transport model. This method realistically depicts the nonlinear characteristics of crowd flow and the formation and propagation patterns of local congestion. Simultaneously, it fully incorporates the spatiotemporal evolution of key parameters such as temperature, visibility, and CO concentration in a fire environment, and establishes a mechanism for quantifying fire risk exposure levels and correcting traffic capacity, making the evacuation simulation results closer to the complex changes under actual fire conditions. Based on this, the method innovatively introduces the direction of dynamic wayfinding signs as a real-time adjustable decision variable. Combining a robust optimization framework that integrates box-type uncertainty sets and cross-budget uncertainty sets, it constructs a multi-objective optimization system centered on cumulative passenger evacuation time, fire risk exposure levels, and the balance of personnel density in key areas. This effectively balances the contradiction between evacuation efficiency and safety, and maintains high performance stability even under the most unfavorable fire disturbance scenarios. Furthermore, by deeply integrating the Zigzag adaptive escape mechanism of the kangaroo escape optimization algorithm with the genetic algorithm, the global search capability and local fine mining capability of high-dimensional discrete combinatorial optimization problems are significantly improved. This ensures the solution efficiency and quality of the algorithm in complex and uncertain environments, thus providing more reliable technical support for real-time, intelligent, and robust evacuation guidance in large and complex subway stations under uncertain fire conditions. It has good engineering application value and practical promotion prospects.

[0187] Based on the above embodiments, Figure 4 This is a schematic diagram illustrating the solution of a robust optimization model provided in an embodiment of this disclosure, specifically including, as follows: Figure 4 The following steps are shown:

[0188] First, all dynamic directional signs within the subway station are discretely encoded, mapping the feasible directions of each sign at each time step to an integer sequence, forming an initial population. Then, a global search phase using a genetic algorithm is initiated, continuously generating new candidate directional sign schemes through roulette wheel selection, two-point crossover, and random mutation operations. For each candidate scheme, its directional decision is fixed and used as input to the outer-layer optimization problem, entering the inner-layer robust subproblem solving process. At this point, based on the nominal risk value, under the constraints of box uncertainty and budget uncertainty, the system searches for the worst-case risk perturbation scenario that results in the worst evacuation performance. The Gurobi solver is used to accurately calculate the maximum risk deviation increment under this scenario and feeds it back into the comprehensive fitness function to evaluate the robustness of the directional scheme in the worst-case scenario. After the GA completes one generation of evolution, individuals with better fitness are selected from the current population, and the kangaroo escape optimization algorithm is further introduced for local augmentation search. The KEO module first calculates the current optimal solution direction vector. Then, it simulates the rapid kangaroo-like behavior by randomly rotating the vector to generate a Zigzag escape direction. A Gaussian perturbation is then applied to achieve a nonlinear jump update, thus refining the perturbation of candidate solutions. After continuous updates, the continuous perturbation results are mapped back to the discrete pointing domain via nearest neighbor projection, ensuring the feasibility of the solution. The updated individuals recalculate their overall fitness and participate in the next generation competition along with their offspring generated by GA. Through the coordinated iteration of global search and local escape enhancement, until the maximum number of iterations or fitness convergence condition is reached, the optimal dynamic guidance sign pointing strategy that balances evacuation efficiency, safety, and balanced population distribution in uncertain fire conditions is finally output.

[0189] Figure 5 This is a schematic diagram of a dynamic wayfinding sign optimization device provided in an embodiment of the present disclosure. The dynamic wayfinding sign optimization device provided in this embodiment can execute the processing flow provided in the dynamic wayfinding sign optimization method embodiment, such as… Figure 5 As shown, the dynamic wayfinding sign optimization device 500 includes:

[0190] The first model construction unit 501 is used to divide the internal evacuation space of the target subway station into multiple cells, and establish the connectivity between multiple cells according to the actual spatial structure of the target subway station, the preset channel connection relationship and the preset evacuation, to construct the spatial discrete model of the target subway station. The cells are used to represent the local evacuation areas inside the subway station.

[0191] The second model construction unit 502 is used to define the number of passengers in each cell within the current time step and the passenger transmission flow between adjacent cells, and to deduce the evacuation process of passengers in multiple cells based on preset supply and demand constraints, and to construct a cell transmission model.

[0192] Model correction unit 503 is used to obtain fire environment parameters at different times in the target subway station and map the fire environment parameters to the cell level to correct the passage capability of each cell in the cell transmission model in the fire environment.

[0193] The third model building unit 504 is used to build a dynamic wayfinding sign optimization model with the objectives of cumulative passenger evacuation time, fire risk exposure level and density balance in key areas. The dynamic wayfinding sign decision variable in the dynamic wayfinding sign optimization model represents the pointing state of each wayfinding sign in the current time step, which is used to control the flow path of passengers between cells.

[0194] Robust optimization unit 505 is used to determine the uncertainties in the fire environment based on fire environment parameters, and embed the uncertainties into the dynamic guidance sign optimization model to form a dynamic guidance sign robust optimization model;

[0195] The strategy update unit 506 is used to solve the robust optimization model of dynamic wayfinding signs and obtain the optimal dynamic wayfinding sign pointing strategy of the target subway station under the uncertain fire environment through iterative update.

[0196] Among them, multiple cells include the first cell and the second cell, and the first cell is the adjacent upstream cell of the second cell.

[0197] Optionally, the second model building unit 502 is used for:

[0198] Calculate the current time step based on the physical property parameters of the cell and the free motion speed of the passenger in the free motion state;

[0199] Define the number of passengers in each cell within the current time step;

[0200] Based on the number of passengers in the first cell, the cell area, and the first preset passage conditions, calculate the supply capacity of the first cell in the time step, where the supply capacity is used to characterize the maximum number of passengers that the first cell outputs to the second cell in the current time step.

[0201] Based on the remaining available space of the second cell, the cell area, and the second preset passage conditions, the receiving capacity of the second cell in the time step is calculated, where the receiving capacity is used to characterize the maximum number of passengers received by the second cell from the first cell in the current time step.

[0202] Based on the free movement speed, preset cell critical density, preset cell congestion density and preset congestion movement speed, the evolution characteristics of the passenger flow in the cell over time during the passenger evacuation process are deduced to obtain the actual cell density.

[0203] Based on the actual density, supply capacity, and receiving capacity of the cells, the passenger transmission flow between adjacent cells in the current time step is defined, and a cell transmission model is constructed.

[0204] Optionally, the second model building unit 502 is used for:

[0205] When the first cell is connected to multiple second cells in the current time step, calculate the evacuation field value corresponding to each second cell. The evacuation field value is used to characterize the evacuation superiority in the downstream direction.

[0206] The allocation ratio of population transmission traffic to each second cell is determined based on the evacuation field value corresponding to each second cell.

[0207] Based on the allocation ratio of each second cell, the passenger transport flow between adjacent cells within the current time step is defined according to the evacuation priority of each second cell.

[0208] Optionally, the model correction unit 503 is used for:

[0209] Obtain fire environment parameters at different times within the target subway station, and map the fire environment parameters to the cell level of the cell transfer model to obtain the fire environment state corresponding to each cell at different time steps.

[0210] Fire environment impact factors are constructed based on fire environment conditions. These factors are used to quantify the inhibitory effect of the fire environment on passenger passage capacity.

[0211] The passage capacity of each cell in a free-flowing state is modified based on the fire environment impact factors to obtain the modified cell transmission model. The modified cell transmission model takes into account the effective passage capacity of cells after the fire environment impact.

[0212] Optionally, the model correction unit 503 is used for:

[0213] The quantized value of visibility at a cell is calculated based on the preset weight coefficient of visibility, the actual value of visibility at a cell in the current time step, and the initial value of visibility at a cell in the current time step. The visibility factor is then calculated based on the quantized value of visibility at a cell. The visibility factor is used to characterize the impairment of crowd mobility caused by visual impairment due to smoke.

[0214] The quantified value of CO concentration at the cell is calculated based on the preset weighting coefficient of CO concentration, the actual value of CO concentration at the cell in the current time step, and the initial value of CO concentration at the cell in the current time step. The CO concentration factor is calculated based on the quantified value of CO concentration at the cell. The CO concentration factor is used to characterize the impact of the cumulative effect of CO exposure on the mobility of the population.

[0215] The quantized value of the temperature at the cell is calculated based on the preset weighting coefficient of the temperature, the actual value of the temperature at the cell in the current time step, and the initial value of the temperature at the cell in the current time step. The temperature factor is calculated based on the quantized value of the temperature at the cell. The temperature factor is used to characterize the thermal radiation damage caused by the high temperature environment and the constraint of the heat resistance limit on the mobility of the population.

[0216] Fire environmental impact factors were constructed based on visibility factor, CO concentration factor, and temperature factor.

[0217] Optionally, the dynamic wayfinding signage optimization device 500 is used for:

[0218] The cumulative passenger dwell time is obtained by summing up the passenger dwell time in each cell during the evacuation process.

[0219] The quantified values ​​of visibility, CO concentration, and temperature at the cell are normalized and superimposed according to their weights to obtain the nominal risk value at the cell in the current time step. The nominal risk value is used to characterize the level of environmental risk that passengers experience during the evacuation process.

[0220] Based on the actual cell density of cells in the key areas of the target subway station at the current time step and the average density of all key areas at the current time step, construct the passenger density variance function of the key areas.

[0221] A dynamic wayfinding sign optimization model is constructed based on the cumulative passenger dwell time, nominal risk value, and passenger density variance function.

[0222] Optionally, the robust optimization unit 505 is used for:

[0223] The fire environment parameters are set as uncertain parameters, which are used to characterize the impact of fire development and environmental disturbances on the evacuation operation of the subway station.

[0224] An uncertainty set is constructed for the uncertain parameter, wherein the uncertainty set is used to describe the fluctuation range of the uncertain parameter within the allowable range, and the uncertainty set includes at least a box uncertainty set and a cross-budget uncertainty set;

[0225] Under the constraints of the uncertainty set, the dynamic guidance sign optimization model is robustly processed to form a dynamic guidance sign robust optimization model, which still satisfies the evacuation operation constraints under the most unfavorable conditions.

[0226] The robust optimization model for dynamic guidance signs is transformed to convert the worst-case conditions it contains into a set of deterministic constraints of equal order.

[0227] Optionally, the robust optimization unit 505 is used for:

[0228] Construct an inner-layer maximization problem function based on the population distribution state of the cells, the uncertain parameters, and the worst-case deviation value;

[0229] Construct the Lagrange function based on the non-negative Lagrange multiplier sum maximization problem function;

[0230] Solve the one-dimensional maximization problem of the Lagrange function with respect to the worst-case deviation value to obtain the optimal value of the worst-case deviation;

[0231] Substituting the optimal value into the maximization problem function, and based on the optimal deviation contribution of all cells and the Lagrange multipliers, the increment of the total evacuation risk value under the worst-case scenario is obtained.

[0232] Substituting the increment of the total evacuation risk value into the pre-constructed minimization problem function yields a dynamic guidance identifier robust optimization model containing a set of deterministic constraints.

[0233] Optionally, the policy update unit 506 is used for:

[0234] The pointing strategy of the dynamic guidance identifier given by the robust optimization model of the dynamic guidance identifier in the current time step is encoded as a discrete decision sequence, which serves as the individual representation of each individual in the genetic algorithm;

[0235] Initialize the genetic algorithm parameters and generate an initial population, and calculate the fitness value of the guidance identifier corresponding to each individual in the initial population according to the deterministic optimization model;

[0236] Based on fitness values, selection, crossover, and mutation operations are performed on individuals in the population to generate candidate dynamic guidance identifiers pointing to strategies.

[0237] During the iterative search in the discrete solution space, the candidate dynamic guidance sign pointing strategies are updated based on the kangaroo escape optimization algorithm to obtain the optimal dynamic guidance sign pointing strategy for the target subway station under the uncertain fire environment.

[0238] Optionally, the policy update unit 506 is used for:

[0239] Select a subset of individuals from the population of the genetic algorithm and use the guidance markers corresponding to these individuals to point to the decision sequence as the initial solution input into the kangaroo escape optimization algorithm;

[0240] Construct a direction vector pointing to the current optimal solution, where the direction vector is used to ensure that the individual has the ability to move to a better region;

[0241] Apply a random rotation to the direction vector and calculate the rotation angle;

[0242] The rotated escape direction is generated based on the unit vector processed by the direction vector and the rotation angle;

[0243] Based on the escape direction and preset jump intensity parameters, a nonlinear offset is achieved in the individual search process by setting a Gaussian perturbation term, and a direction guidance term between the optimal individual and the current individual is introduced to prompt the individual to jump towards the current optimal solution direction for update.

[0244] After performing continuous perturbation updates on individuals, the updated individuals are mapped back to the preset discrete domain through nearest neighbor projection to ensure that the updated perturbation maintains the escape search characteristics while satisfying the feasibility constraints of the dynamic guidance label pointing strategy.

[0245] The evacuation efficiency and safety performance of candidate dynamic guidance marker pointing strategies are comprehensively evaluated. Based on the comprehensive evaluation results, the population is selected and updated to continue the next iteration until the termination condition is met, and the iteration stops to obtain the optimal dynamic guidance marker pointing strategy.

[0246] Figure 5 The dynamic guidance sign optimization device shown in the embodiment can be used to execute the technical solution of the above method embodiment. Its implementation principle and technical effect are similar, and will not be repeated here.

[0247] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. See below for details. Figure 6 The diagram illustrates a structural schematic suitable for implementing the electronic device 600 in the embodiments of this disclosure. The electronic device 600 in the embodiments of this disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), wearable electronic devices, etc., as well as fixed terminals such as digital TVs, desktop computers, smart home devices, etc. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0248] like Figure 6 As shown, the electronic device 600 may include a processing unit 601 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603 to implement the dynamic navigation identifier optimization method as described in the embodiments of this disclosure. The RAM 603 also stores various programs and data required for the operation of the electronic device 600. The processing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0249] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0250] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts, thereby implementing the dynamic guidance identification optimization method as described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of embodiments of this disclosure.

[0251] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0252] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0253] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0254] Optionally, when one or more of the above-described procedures are executed by the electronic device, the electronic device may also perform other steps described in the above embodiments.

[0255] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0256] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0257] The units described in the embodiments of this disclosure can be implemented in software or hardware. The names of the units are not, in some cases, intended to limit the specific unit.

[0258] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0259] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0260] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or gateway that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or gateway. Without further limitations, an element defined by the phrase "comprising a dynamic guidance identifier optimization" does not exclude the presence of other identical elements in the process, method, article, or gateway that includes said element.

[0261] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing dynamic wayfinding signs, characterized in that, include: The internal evacuation space of the target subway station is divided into multiple cells, and the connectivity between the multiple cells is established according to the actual spatial structure of the target subway station, the preset channel connection relationship and the preset evacuation. The spatial discrete model of the target subway station is constructed, wherein the cells are used to represent local evacuation areas inside the subway station. Define the number of passengers in each cell within the current time step and the passenger transmission flow between adjacent cells, and deduce the evacuation process of passengers in the multiple cells based on preset supply and demand constraints to construct a cell transmission model; The fire environment parameters at different times within the target subway station are obtained, and the fire environment parameters are mapped to the cell level to correct the passage capability of each cell in the cell transmission model under the fire environment. A dynamic wayfinding sign optimization model is constructed with the objectives of cumulative passenger evacuation time, fire risk exposure level, and density balance in key areas. The dynamic wayfinding sign decision variable in the dynamic wayfinding sign optimization model represents the pointing state of each wayfinding sign in the current time step, which is used to control the flow path of passengers between cells. Based on the fire environment parameters, the uncertainties under the fire environment are determined, and these uncertainties are embedded into the dynamic wayfinding sign optimization model to form a robust dynamic wayfinding sign optimization model. This includes: setting the fire environment parameters as uncertain parameters, wherein the uncertain parameters characterize the impact of fire development and environmental disturbances on the evacuation operation of the subway station; constructing an uncertainty set for the uncertain parameters, wherein the uncertainty set describes the fluctuation range of the uncertain parameters within an allowable range, and the uncertainty set includes at least a box-type uncertainty set and a cross-calculation uncertainty set; and robustening the dynamic wayfinding sign optimization model under the constraints of the uncertainty set to form a robust dynamic wayfinding sign optimization model, which aims to achieve the best possible outcome in the least uncertain environment. Under favorable conditions, the evacuation operation constraints are still met; an inner-layer maximization problem function is constructed based on the population distribution state of the cells, the uncertain parameters, and the worst-case deviation value; a Lagrange function is constructed based on the non-negative Lagrange multipliers and the maximization problem function; a one-dimensional maximization problem of the Lagrange function with respect to the worst-case deviation value is solved to obtain the optimal value of the worst-case deviation; the optimal value is substituted into the maximization problem function, and the total evacuation risk increment under the worst-case condition is obtained based on the optimal deviation contribution of all cells and the Lagrange multiplier terms; the total evacuation risk increment is substituted into the pre-constructed minimization problem function to obtain the dynamic guidance sign robust optimization model containing a set of deterministic constraints; Solving the robust optimization model for dynamic wayfinding signs and obtaining the optimal dynamic wayfinding sign direction strategy for the target subway station under uncertain fire conditions through iterative updates includes: encoding the direction strategy of the dynamic wayfinding signs given by the robust optimization model within the current time step into a discrete decision sequence, which serves as the individual representation of each individual in the genetic algorithm; initializing the genetic algorithm parameters and generating an initial population, and calculating the fitness value of the wayfinding sign corresponding to each individual in the initial population according to the deterministic optimization model; performing selection, crossover, and mutation operations on the individuals in the population based on the fitness value to generate candidate dynamic wayfinding sign direction strategies; and updating the candidate dynamic wayfinding sign direction strategies based on the kangaroo escape optimization algorithm during the iterative search in the discrete solution space to obtain the optimal dynamic wayfinding sign direction strategy for the target subway station under uncertain fire conditions.

2. The method according to claim 1, characterized in that, The plurality of cells includes a first cell and a second cell, wherein the first cell is the adjacent upstream cell of the second cell. The method defines the number of passengers in each cell within the current time step and the passenger transport flow between adjacent cells, and constructs a cell transport model based on preset supply and demand constraints to deduce the passenger evacuation process within the plurality of cells, including: The current time step is calculated based on the physical property parameters of the cell and the free movement speed of the passenger in the free movement state; Define the number of passengers in each cell within the current time step; Based on the number of passengers in the first cell, the cell area, and the first preset passage conditions, the supply capacity of the first cell in the time step is calculated, wherein the supply capacity is used to characterize the maximum number of passengers output by the first cell to the second cell in the current time step. Based on the remaining available space of the second cell, the cell area, and the second preset passage conditions, the receiving capacity of the second cell in the time step is calculated, wherein the receiving capacity is used to characterize the maximum number of passengers received by the second cell from the first cell in the current time step. Based on the free movement speed, preset cell critical density, preset cell congestion density and preset congestion movement speed, the evolution characteristics of the passenger flow in the cell over time during the passenger evacuation process are deduced to obtain the actual cell density. Based on the actual cell density, the supply capacity, and the receiving capacity, the passenger transmission flow between adjacent cells in the current time step is defined, and a cell transmission model is constructed.

3. The method according to claim 2, characterized in that, Define the passenger transport flow between adjacent cells within the current time step, including: When the first cell is connected to multiple second cells in the current time step, the evacuation field value corresponding to each second cell is calculated, wherein the evacuation field value is used to characterize the evacuation superiority in the downstream direction. The allocation ratio of the population transmission traffic for each second cell is determined based on the evacuation field value corresponding to each second cell. Based on the allocation ratio of each second cell, the passenger transport flow between adjacent cells in the current time step is defined according to the evacuation priority of each second cell.

4. The method according to claim 1, characterized in that, The step of acquiring fire environment parameters at different times within the target subway station and mapping these parameters to the cell level to correct the mobility of each cell in the cell transmission model within the fire environment includes: The fire environment parameters at different times within the target subway station are obtained, and the fire environment parameters are mapped to the cell level of the cell transmission model to obtain the fire environment state corresponding to each cell at different time steps. Fire environment impact factors are constructed based on the fire environment status, wherein the fire environment impact factors are used to quantify the inhibitory effect of the fire environment on passenger passage capacity. The passage capacity of each cell in a free-flowing state is modified based on the fire environment impact factor to obtain the modified cell transmission model, wherein the modified cell transmission model takes into account the effective passage capacity of cells after the fire environment impact.

5. The method according to claim 4, characterized in that, The construction of fire environment impact factors based on the fire environment state includes: The quantized value of visibility at the cell is calculated based on the preset weight coefficient of visibility, the actual value of visibility at the cell in the current time step, and the initial value of visibility at the cell in the current time step. A visibility factor is then calculated based on the quantized value of visibility at the cell. The visibility factor is used to characterize the impairment of crowd mobility caused by visual impairment due to smoke. The quantified value of the CO concentration at the cell is calculated based on the preset weighting coefficient of the CO concentration, the actual value of the CO concentration at the cell in the current time step, and the initial value of the CO concentration at the cell in the current time step. The CO concentration factor is calculated based on the quantified value of the CO concentration at the cell. The CO concentration factor is used to characterize the impact of the cumulative effect of CO exposure on the mobility of the population. The quantized value of the temperature at the cell is calculated based on the preset weighting coefficient of the temperature, the actual value of the temperature at the cell in the current time step, and the initial value of the temperature at the cell in the current time step. The temperature factor is calculated based on the quantized value of the temperature at the cell. The temperature factor is used to characterize the thermal radiation damage caused by the high temperature environment and the constraint of the heat resistance limit on the mobility of the population. Fire environmental impact factors are constructed based on the visibility factor, the CO concentration factor, and the temperature factor.

6. The method according to claim 1, characterized in that, The construction of the dynamic wayfinding sign optimization model, which aims to optimize passenger cumulative evacuation time, fire risk exposure level, and density balance in key areas, includes: The cumulative passenger dwell time is obtained by summing the passenger dwell time in each cell during the evacuation process; The quantified values ​​of visibility, CO concentration, and temperature at the cell are normalized and superimposed according to weights to obtain the nominal risk value at the cell in the current time step, wherein the nominal risk value is used to characterize the level of environmental risk borne by passengers during evacuation. Based on the actual cell density of cells in the key areas of the target subway station in the current time step and the average density of all key areas in the current time step, construct the passenger density variance function of the key areas. A dynamic wayfinding sign optimization model is constructed based on the cumulative passenger dwell time, the nominal risk value, and the passenger density variance function.

7. The method according to claim 1, characterized in that, During the iterative search in the discrete solution space, the candidate dynamic wayfinding sign pointing strategies are updated based on the kangaroo escape optimization algorithm to obtain the optimal dynamic wayfinding sign pointing strategy for the target subway station under uncertain fire conditions, including: Select a subset of individuals from the population of the genetic algorithm, and use the guidance markers corresponding to these individuals to point to the decision sequence as the initial solution input into the kangaroo escape optimization algorithm; Construct a direction vector pointing to the current optimal solution, wherein the direction vector is used to ensure that the individual has the ability to move to a better region; Apply a random rotation to the direction vector and calculate the rotation angle; The rotated escape direction is generated based on the unit vector disposed with the direction vector and the rotation angle; Based on the escape direction and the preset jump intensity parameters, a nonlinear offset is achieved in the individual search process by setting a Gaussian perturbation term, and a direction guidance term between the optimal individual and the current individual is introduced to prompt the individual to jump towards the current optimal solution direction for update. After performing continuous perturbation updates on individuals, the updated individuals are mapped back to the preset discrete domain through nearest neighbor projection to ensure that the updated perturbation maintains the escape search characteristics while satisfying the feasibility constraints of the dynamic guidance label pointing strategy. The evacuation efficiency and safety performance of the candidate dynamic guidance marker pointing strategies are comprehensively evaluated, and the population is selected and updated based on the comprehensive evaluation results to continue the next iteration until the termination condition is met and the iteration stops, thus obtaining the optimal dynamic guidance marker pointing strategy.

Images