A Method for Constructing and Simulating a Crowd Flow Simulation Model Based on Agents

Abstract

This invention provides an agent-based method for constructing and simulating pedestrian flow models. The method includes: constructing a regional model, where a scenic area is uniformly divided into several grids; setting basic attributes, carrying capacity attributes, and reachability attributes for each grid to construct grid attributes; all grid attributes collectively forming the regional model of the scenic area; constructing an entity model, where basic attributes, tour attributes, and status attributes for each tourist are set to form a tourist entity model; constructing a behavior model, where tourist behavior is formed by setting decisions regarding attraction stay, attraction selection, movement speed, and exit based on the entity model; and predicting pedestrian flow, where several tourist information sets are used to generate several entity models. These entity models are then fed into the behavior model and the regional model to predict the movement path of each tourist within the regional model, thereby predicting changes in pedestrian flow within different scenic areas. This invention can predict changes in pedestrian flow within different attractions within a scenic area during a specific time period, thus enabling advance prediction of the overall tourist carrying capacity of the scenic area.

Description

Technical Field

[0001] This invention relates to the field of crowd flow simulation technology, specifically to an agent-based crowd flow simulation model construction and simulation method. Background Technology

[0002] With the booming development of global tourism, more and more people are eager to visit various cultural heritage sites to explore the mysteries of history and experience the charm of ancient cultures. However, cultural heritage sites often face the dilemma of overcrowding. Large numbers of tourists gather in relatively small spaces with limited carrying capacity, putting unbearable strain on the environment of cultural relics. Unconscious physical contact and frequent trampling by tourists cause direct wear and damage to the physical structure of cultural relics and historical sites. Stone slabs become dented and cracked due to excessive trampling, and the wooden carvings on the doors and windows of ancient buildings become blurred from frequent touching. These phenomena of damage are commonplace.

[0003] Existing technological approaches primarily focus on two key aspects: visitor capacity limits and visitor flow management strategies. Visitor capacity limits are determined through scientific calculations, taking into account factors such as the spatial scale of the cultural heritage site, the fragility of the artifacts, and the capacity of supporting facilities, to establish an upper limit for the number of visitors that can be accommodated. When the number of visitors approaches the threshold, ticket sales are immediately suspended or flow control measures are implemented to prevent overcrowding. Visitor flow management strategies utilize various online and offline channels to proactively send visitors information on visitor flow at different times and in different areas, guiding them to visit during off-peak hours. This ensures a good visitor experience while minimizing the risk of damage to precious cultural relics. Summary of the Invention

[0004] In view of this, the problem to be solved by the present invention is to provide a method for constructing and simulating a person flow simulation model based on an agent.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for constructing a pedestrian flow simulation model based on an agent, comprising: constructing a regional model of a scenic area, dividing the scenic area into several grids evenly, setting the basic attributes, carrying capacity attributes and reachability attributes of the grids to construct grid attributes, and all grid attributes together constituting the regional model of the scenic area;

[0006] Construct a physical model of the tourist, and set the tourist's basic attributes, tour attributes, and status attributes to form the tourist's physical model;

[0007] Construct a tourist behavior model, and based on the entity model, set the tourist's attraction stay decision, attraction selection decision, tourist movement speed, and exit decision to form the tourist's behavior model;

[0008] The pedestrian flow prediction method involves setting up several visitor information to generate several entity models. These entity models are then fed into a behavior model and a regional model to predict the movement path of each visitor within the regional model, thereby predicting changes in pedestrian flow in different scenic areas.

[0009] Furthermore, the basic attributes of the grid record the grid's address information and attractiveness level; the grid's carrying capacity attribute records the maximum number of tourists the grid can carry and the tourist movement speed corresponding to different tourist carrying capacities; and the grid's reachability attribute records whether the grid can carry tourists.

[0010] Furthermore, the basic attributes of the tourist include the tourist's age information, location information, entry time and exit time of the scenic area, and the tourist's tour attributes include the tourist's energy attributes and interest attributes. The energy attributes include the tourist's energy level while visiting the scenic area, and the interest attributes include attraction interest level, attraction interest decay level, attraction interest recovery level, minimum attraction interest value, and maximum attraction interest value. The tourist's status attributes include information on attractions not visited by the tourist, information on attractions visited by the tourist, an indicator of whether the tourist has moved to the target location, and an indicator of whether the tourist has left the scenic area.

[0011] Furthermore, the decision to stay at a tourist attraction is based on two conditions: First, when a tourist's interest in the current attraction drops below its minimum interest level, the tourist will tend to leave; second, if the tourist's stay at the current attraction exceeds a reasonable duration, it will also trigger a decision to leave. The formula for updating the interest level for the current attraction is:

[0012] SIV i，current,t =(1-IDR) i,current (1-ADT) i,current )×(SIV i，current，t-1 -MIV i，current ),

[0013] Among them, SIV i,current,t and SIV i,current,t-1 IDR represents the interest value of the i-th tourist at time t and time t-1 for the current attraction. i,current ADT represents the decay rate of the interest value of the i-th tourist towards the current attraction. i,current MIV represents the recovery rate of the interest value of the i-th tourist to the attraction they are visiting. i,current This represents the minimum interest value of the i-th tourist for the current attraction.

[0014] Furthermore, the attraction selection decision follows a priority rule: among attractions not yet visited, tourists evaluate them based on their current interest level and select the attraction with the highest interest. The formula for updating the interest level for unvisited attractions is as follows:

[0015] SIV i，j，t =(IRR) i，j –1)(SIV i，j，t-1 -MAV i，j ),

[0016] Among them, SIV i,j,t and SIV i,j,t-1 IRR represents the interest value of the i-th tourist for the j-th attraction at time t and time t-1. i,j MAV represents the recovery rate of interest of the i-th tourist towards the j-th unvisited attraction. i,j This represents the maximum interest of the i-th tourist in the j-th attraction.

[0017] Furthermore, the formula for the tourist's movement speed is:

[0018]

[0019] Among them, SR x,y,t SC represents the velocity ratio at time t within the grid corresponding to the x and y coordinate positions. x,y,t This represents the ratio of the total number of people in the grid corresponding to the x and y coordinate positions to the maximum capacity.

[0020] Furthermore, the exit decision includes obtaining the tourist's status attributes, determining whether the list of unvisited attractions is empty; if yes, leaving the scenic area; if no, obtaining the movement speed and the list of unvisited attractions, and selecting the attraction with the highest interest from the list of unvisited attractions for the next visit.

[0021] If the time to reach the next attraction exceeds the park's closing time, leave the park; otherwise, determine the visitor's current energy value using an energy calculation formula. If the current energy value is below a threshold, leave the park; otherwise, continue visiting.

[0022] Furthermore, the energy calculation formula is as follows:

[0023]

[0024] Among them, E i,t and E i,t-1 EDR represents the energy value of the i-th tourist at time t and time t-1. i E represents the energy decay rate of the i-th tourist. mini This represents the minimum energy value for the i-th visitor.

[0025] Furthermore, the pedestrian flow prediction includes optimizing the parameters of the region model and the entity model using Optuna. The Optuna optimization includes determining the optimization result using a loss function; the loss function is:

[0026]

[0027] Where N represents the number of attractions involved in the data, Y i Let represent the actual visit time distribution for the i-th attraction. The simulation program obtained the visit time distribution for the i-th attraction. Let λ represent the actual average visit time for the entire scenic area, T represent the simulated average visit time for the entire scenic area, and λ represent the importance of the total average visit time.

[0028] A method for simulating pedestrian flow includes initializing a physical model of tourists and a regional model of a scenic area; predicting whether tourists will continue their visit based on an exit decision; if not, determining the optimal exit, generating a navigation path for the scenic area and moving there; if yes, determining whether tourists are entering the grid for the first time.

[0029] If it is the first time entering the grid area, the system predicts the tourist's chosen attractions based on the attraction selection decision, generates a navigation path, and moves accordingly. If it is not the first time entering the grid area, the system determines whether the tourist has reached the target attraction. If not, the tourist moves along the existing navigation path. If yes, the system determines whether the tourist stays to visit the attraction based on the attraction dwell decision.

[0030] If a tourist stays to visit, the system predicts the tourist's movement path based on the tourist's movement speed, generates a navigation path, and moves accordingly. If a tourist does not stay to visit, the system predicts the tourist's next destination based on the tourist's visit decision, generates a navigation path, and moves accordingly.

[0031] Update visitor status and capabilities, and then proceed to the step of predicting whether the visitor will continue their tour based on exit decisions.

[0032] The advantages and positive effects of the present invention are:

[0033] (1) By constructing regional models, entity models and behavior models respectively, and constructing entity models for each tourist based on entity models and tourist information, each tourist's entity model is fed into the behavior model and regional model to predict the visit path of each tourist in the scenic area. This allows for efficient and accurate prediction of the flow of people in different areas of the scenic area, so as to predict the tourist carrying capacity of the scenic area in advance and reduce the probability of damage to attractions in the scenic area due to excessive flow of people.

[0034] (2) Improve the accuracy of human flow prediction by setting loss functions to optimize the parameters within the regional model and the entity model. Attached Figure Description

[0035] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0036] Figure 1This is a diagram illustrating the construction process of an Agent-based crowd flow simulation model according to the present invention;

[0037] Figure 2 This is an overall flowchart of the construction and simulation method of a person flow simulation model based on Agent according to the present invention. Detailed Implementation

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0040] This invention provides an agent-based crowd flow simulation model. The agent-based modeling (ABM) method simulates the behavior of each agent with autonomous attributes and decision-making capabilities at the micro level, and is widely used in the field of crowd flow simulation.

[0041] like Figure 1 As shown, this includes constructing a regional model of the scenic area. The scenic area is evenly divided into several grids, and the basic attributes, carrying capacity attributes, and reachability attributes of the grids are set to construct the grid attributes. All grid attributes together constitute the regional model of the scenic area.

[0042] The regional model transforms complex real-world scenes into virtual environments for subsequent pedestrian flow simulation. The generation process of the regional model includes: acquiring a two-dimensional map of the scenic area and dividing it into several grids, setting the basic attributes, carrying capacity attributes, and reachability attributes of each grid, thus forming the regional model of the scenic area.

[0043] The formula for mesh properties is:

[0044] SSM=(S B ,S C ,S R ),

[0045] Wherein: S B S represents the basic attribute. C S represents the carrying capacity attribute. RThis indicates a reachable property.

[0046] The basic attributes record the grid's address information and attractiveness level. The address information is the grid's coordinate range, and whether a visitor enters the grid is determined by whether they enter the coordinate range. The scenic area includes several attractions, each of which consists of several grids.

[0047] One embodiment of this application is as follows: the grid accessible to tourists includes a regular grid and a scenic spot grid. The regular grid is the grid that tourists pass through when moving, and the scenic spot grid is the grid that tourists stay in when visiting scenic spots. Each scenic spot is assigned a different attraction value.

[0048] The carrying capacity attribute records the maximum number of visitors a grid can accommodate and the corresponding visitor movement speed for different carrying capacities. The more visitors within a grid, the slower their movement, reflecting the impact of visitor density on mobility. The carrying capacity of a grid is related to the size of the grid area and the attributes of the area where the grid is located; for example, indoor and outdoor grids have different carrying capacities. In actual settings, the visitor carrying capacity is determined based on the grid's location and the grade of the cultural relics within the grid.

[0049] The reachability attribute records whether the grid is open to visitors. The reachability of closed areas within the scenic area is 0, while the reachability of open areas is 1. Recording the reachability attribute of a grid allows each grid to correspond to a fixed location within the scenic area, improving the realism of the area model.

[0050] One embodiment of this application is: the reachability attribute is adjustable data, and the manager can adjust the flow of people in a timely manner according to the reachability attribute of each grid area to reduce the probability of tourist congestion.

[0051] The regional model includes grid attribute data for all grids. Since the grids are divided according to the real scene, the grid attributes are all set manually based on the actual connectivity of the scenic area, the opening status of attractions, and the relationship between grids and attractions, ensuring the consistency between the regional model and the actual scene.

[0052] Construct a physical model of the tourist, defining the tourist's basic attributes, tour attributes, and status attributes to form the tourist's physical model.

[0053] The formula for the solid model is:

[0054] TEM = (T BA T VA T SA ),

[0055] Among them, T BA T represents the basic attributes of tourists. VA Indicates the visitor's tour attributes, T SA This indicates the visitor's status attributes.

[0056] The basic attributes include the visitor's age, location, entry time, and exit time. Before the visitor flow simulation, each visitor's age is manually set; different ages correspond to different initial energy levels (energy for exploring the area). Location information records the visitor's position within the scenic area and is continuously updated as the simulation progresses. Entry time is obtained by sampling the visitor entry curve. The moment a visitor leaves the scenic area during the simulation is defined as their exit time.

[0057] The formula for expressing tour attributes is:

[0058] T VA =(T VAE T VAI )

[0059] Among them, T VAE T represents the energy attribute. VAI This section describes the interest attributes. Tour attributes include the visitor's current energy value and minimum energy value. The visitor's energy value can be obtained in advance through experimentation or set based on experience. The longer the visit, the lower the energy value; the minimum energy value can be used to control whether a visitor leaves the scenic area. Interest attributes can predict the probability of a visitor actually visiting a attraction. Interest attributes include five categories: attraction interest level, attraction interest decay rate, attraction interest recovery rate, attraction minimum interest value, and attraction maximum interest value. All five categories are manually set based on the visitor's age group and the attraction's online popularity.

[0060] The status attributes record the list of attractions not visited by the tourist, the list of attractions visited by the tourist, an indicator of whether the tourist has moved to the target location, and an indicator of whether the tourist has left the scenic area.

[0061] Construct a tourist behavior model, and based on the entity model, set the tourist's decisions on staying at attractions, choosing attractions, moving speed, and exiting to form the tourist's behavior model.

[0062] Attraction Detention Decision is used to predict whether tourists will stay at the current attraction. The relevant formula for attraction detention decision is:

[0063] SIV i，current,t =(1-IDR) i,current (1-ADT) i，current )×(SIV i，current,t-1 -MIV i，current ),

[0064] Among them, SIV i,current,t and SIV i,current,t-1 IDR represents the interest value of the i-th tourist at time t and time t-1 for the current attraction. i,current ADT represents the decay rate of the interest value of the i-th tourist towards the current attraction. i,currentMIV represents the recovery rate of the interest value of the i-th tourist to the attraction they are visiting. i,current This represents the minimum interest value of the i-th tourist for the current attraction.

[0065] The attraction selection decision is used to predict the next attraction a tourist wants to visit. The relevant formula for attraction selection decision is:

[0066] SIV i，j，t =(IRR) i，j -1)(SIV i，j，t-1 -MAV i，j ),

[0067] Among them, SIV i,j,t and SIV i,j,t-1 IRR represents the interest value of the i-th tourist for the j-th attraction at time t and time t-1. i,j MAV represents the recovery rate of interest of the i-th tourist towards the j-th unvisited attraction. i,j This represents the maximum interest of the i-th tourist in the j-th attraction.

[0068] Tourist movement speed is used to predict the speed at which tourists move within different grid cells. The formula for tourist movement speed is:

[0069]

[0070] Among them, SR x,y,t SC represents the velocity ratio at time t within the grid corresponding to the x and y coordinate positions. x,y,t This represents the ratio of the total number of people in the grid corresponding to the x and y coordinate positions to the maximum carrying capacity. In one embodiment of this application, the maximum moving speed of the tourists is 4 m / s, and the product of the speed ratio and the maximum moving speed is the moving speed of the tourists within the reachable grid at time t.

[0071] Exit decision is used to predict whether tourists will choose to leave the scenic area. The scenic area exit decision includes obtaining the tourist's state attributes from the tourist's entity model, and judging whether the list of unvisited attractions is empty based on the real-time updated state attributes. If it is, it means that the tourist has visited all attractions and will choose to leave the scenic area. If not, it means that there are attractions in the scenic area that the tourist has not visited (the list of unvisited attractions is not empty), and the tourist with the highest interest is selected from the list of unvisited attractions for the next visit.

[0072] The decision to exit the scenic area includes obtaining the current grid location of the tourist, generating a movement path for the tourist to move to the nearest exit based on the current location, determining the movement speed based on the upstream visitor flow density of the path, and determining whether the time it takes for the tourist to move to the nearest exit exceeds the scenic area's closing time. If yes, the tourist leaves the scenic area; otherwise, the tourist continues to visit.

[0073] The decision to exit a scenic area involves obtaining the total time tourists spend in the area and determining whether their current energy value is below the energy threshold using an energy calculation formula. If yes, the tourist leaves the scenic area; otherwise, they continue their visit.

[0074] Energy calculation is used to determine changes in tourists' physical strength. The energy calculation formula is as follows:

[0075]

[0076] Among them, E i,t and E i,t-1 EDR represents the energy value of the i-th tourist at time t and time t-1. i E represents the energy decay rate of the i-th tourist. mini This represents the minimum energy value for the i-th visitor.

[0077] The pedestrian flow prediction method involves setting up several visitor information to generate several entity models. These entity models are then fed into a behavior model and a regional model to predict the movement path of each visitor within the regional model, thereby predicting changes in pedestrian flow in different scenic areas.

[0078] When predicting visitor flow, parameters are set in the regional model and the entity model, and some visitor information is set. The visitor information is input into the entity model to generate an entity model for each visitor. The entity model is then input into the behavior model and the regional model to predict the movement path of each visitor within the regional model and determine the changes in visitor flow at each attraction in the scenic area. This allows for the prediction of the maximum capacity of the scenic area in advance, avoiding the situation where the scenic area receives too many visitors, which could lead to damage to cultural relics within the scenic area.

[0079] Because the regional and entity models contain a large number of manually set parameters, there is a significant gap between the model output and the actual results. Optuna's deep learning hyperparameter tuning method can be used to optimize these parameters and improve prediction accuracy. Optuna hyperparameter tuning involves obtaining a small sample of real visitor visitor path information and adjusting parameters based on this information. Since using a small sample of real visitor data for hyperparameter tuning can lead to errors if the visitor sample is too limited, a loss function is defined. The accuracy of parameter optimization is judged based on the loss function value. If the loss function value meets a threshold, the reliability of the hyperparameter tuning is high, and the results can be adopted; otherwise, the results are not adopted.

[0080] The loss function is:

[0081]

[0082] Where N represents the number of attractions involved in the data, Y i Let represent the actual visit time distribution for the i-th attraction. The simulation program obtained the visit time distribution for the i-th attraction. Let T be the actual average visit time for the entire scenic area, T be the simulated average visit time for the entire scenic area, and λ represent the importance of the total average visit time. In one embodiment of this application, λ is set to 2.

[0083] A method for simulating pedestrian flow, such as Figure 2 As shown, this includes initializing the entity model of tourists and the regional model of the scenic area. The initialization entity model assigns corresponding initial attribute values ​​to each tourist, and the initialization regional model calibrates the attributes of different grids to facilitate subsequent prediction.

[0084] Obtain the flow of people in all grids at time i, and determine the movement speed of tourists in all grids.

[0085] Based on the exit decision prediction, it is determined whether the tourist will continue to visit the scenic area. If the prediction is yes, it means the tourist will continue to visit the scenic area. The current location of the tourist is obtained, and it is determined whether the tourist has entered the grid for the first time. If yes, it means the tourist has arrived at the location for the first time. The next attraction to be visited by the tourist is predicted through the attraction selection decision, and the tourist's navigation path is obtained through BFS. The tourist moves along the navigation path. If no, it means the tourist has entered the grid for the first time. The tourist's state attributes are obtained to determine whether the tourist has moved to the target location, so as to determine whether the tourist has reached the target attraction. If no, it means the tourist is heading to the attraction to be visited along the generated navigation path. The generated navigation path is obtained, and the tourist continues to move along the navigation path. If no, it means the tourist is not in a moving state. The attraction stay decision prediction is determined whether the user will stay and visit the attraction. If yes, a navigation path to visit the attraction is generated based on the attraction characteristics and BFS, and the tourist moves along the navigation path. If no, the next attraction to be visited by the tourist is predicted through the attraction selection decision, and the tourist's navigation path is obtained through BFS. The tourist moves along the navigation path. At time i+1, update the state and energy of all tourists in the prediction model, execute the exit decision, and then jump to the step of determining whether the tourist should continue to visit the scenic area.

[0086] Based on the scenic area exit decision prediction, the system determines whether a tourist will continue to visit the scenic area. If the prediction result is negative, it indicates that the tourist is about to leave the scenic area. The system obtains the tourist's current location information and determines whether the tourist has completely left the scenic area. If yes, it proceeds to the step of determining whether the tourist will continue to visit the scenic area. If no, it continues to determine whether the tourist has determined the best exit. If no, it indicates that the action of leaving the scenic area has been triggered for the first time. The system obtains the tourist's current location information to determine the nearest best exit and obtains the tourist's navigation path using BFS. The tourist moves along the navigation path. If no, it indicates that this is not the first time the action of leaving the scenic area has been triggered. The system obtains the navigation route generated at the first trigger and the tourist moves along the navigation path. Before time i+1, the system obtains the tourist's state attribute and determines whether the tourist has completely left the scenic area (whether the tourist has a departure marker). If yes, it indicates that the tourist has left the scenic area, and the tourist path prediction ends. If no, it indicates that the tourist is still in the scenic area. At time i+1, the tourist's state and energy are updated, and the system proceeds to the step of determining whether the tourist will continue to visit the scenic area.

[0087] The prediction model iterates through the behavior models of all tourists at time i to predict the movement paths of each tourist individually. At time i+1, it updates the tourists' location information, visit status information, energy information, etc., and continues to iterate through the behavior models of all tourists and predict the movement paths of tourists until all tourists have left the scenic area.

[0088] By artificially setting the age combination of tourists and the time curve of tourists entering the scenic area, and then using the flow prediction model and prediction method, the flow changes of tourists in each attraction in the scenic area (flow change curve) and the length of tourists' visit are predicted in order to predict the tourist carrying capacity range of the scenic area.

[0089] One embodiment of this application is: when some attractions in a scenic area are closed for maintenance, the model can predict the actual tourist carrying capacity of the scenic area in advance, thereby reducing the probability that tourists will buy tickets but will not be able to enter the scenic area or will not be able to enjoy themselves in the scenic area. It can also avoid the probability of damage to cultural relics due to too many tourists in the scenic area.

[0090] The embodiments of the present invention are described in detail above, but the contents described are only preferred embodiments of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of this patent.

Claims

1. An agent-based pedestrian flow simulation model, characterized in that, include, Construct a regional model of the scenic area. The scenic area is evenly divided into several grids. Set the basic attributes, carrying capacity attributes, and reachability attributes of the grids to construct the grid attributes. All grid attributes together constitute the regional model of the scenic area. Construct a physical model of the tourist, and set the tourist's basic attributes, tour attributes, and status attributes to form the tourist's physical model; Construct a tourist behavior model, and based on the entity model, set the tourist's attraction stay decision, attraction selection decision, tourist movement speed, and exit decision to form the tourist's behavior model; The pedestrian flow prediction method involves setting up several tourist information to generate several entity models. These entity models are then fed into a behavior model and a regional model to predict the movement path of each tourist within the regional model, thereby predicting changes in pedestrian flow in different scenic areas. The decision to stay at a tourist attraction is based on two conditions: First, when a tourist's interest in the current attraction drops below its minimum interest level, the tourist will tend to leave; second, if the tourist's stay at the current attraction exceeds a reasonable duration, it will also trigger a decision to leave. The formula for updating the interest level for the current attraction is as follows: USE i,current,t =(1-IDR i，current )(1-ADT i，current )×(USE i，current,t-1 -CAT i,current ), Among them, SIV i,current,t and SIV i,current,t-1 IDR represents the interest value of the i-th tourist at time t and time t-1 for the current attraction. i,current ADT represents the decay rate of the interest value of the i-th tourist towards the current attraction. i,current MIV represents the recovery rate of the interest value of the i-th tourist to the attraction they are visiting. i,current This represents the minimum interest value of the i-th tourist for the current attraction; The attraction selection decision follows a priority rule: among attractions not yet visited, tourists evaluate them based on their current interest level and select the attraction with the highest interest. The formula for updating the interest level for unvisited attractions is as follows: USE ij，t =(IRR i，j –1)(USE i，j,t-1 -CAT i，j ), Among them, SIV i,j,t and SIV i,j,t-1 IRR represents the interest value of the i-th tourist for the j-th attraction at time t and time t-1. i,j MAV represents the recovery rate of interest of the i-th tourist towards the j-th unvisited attraction. i,j This represents the maximum interest of the i-th tourist in the j-th attraction; The formula for the tourist's movement speed is: Among them, SR x,y,t SC represents the velocity ratio at time t within the grid corresponding to the x and y coordinate positions. x,y,t This represents the ratio of the total number of people in the grid corresponding to the x and y coordinate positions to the maximum capacity; The exit decision includes obtaining the tourist's status attributes, determining whether the list of unvisited attractions is empty, and if so, leaving the scenic area; if not, selecting the attraction with the highest interest from the list of unvisited attractions for the next visit. The exit decision includes determining whether the tourist's current energy value is lower than the energy threshold using an energy calculation formula; if yes, the tourist leaves the scenic area; if no, the tourist continues to visit. The exit decision includes obtaining the tourist's movement speed and determining whether the time it takes for the tourist to move to the nearest exit exceeds the park's closing time. If yes, the tourist leaves the park; if no, the tourist continues to visit. The formula for calculating tourist energy renewal is as follows: Among them, E i,t and E i,t-1 EDR represents the energy value of the i-th tourist at time t and time t-1. i E represents the energy decay rate of the i-th tourist. mini This represents the minimum energy value for the i-th visitor; The pedestrian flow prediction includes optimizing the parameters of the region model and the entity model using Optuna. The Optuna optimization includes determining the optimization result using a loss function; the loss function is: Where N represents the number of attractions involved in the data, Y i Let represent the actual visit time distribution for the i-th attraction. The simulation program obtained the visit time distribution for the i-th attraction. Let λ represent the actual average visit time for the entire scenic area, T represent the simulated average visit time for the entire scenic area, and λ represent the importance of the total average visit time.

2. The agent-based crowd flow simulation model according to claim 1, characterized in that, The basic attributes of the grid record the grid's address information and attractiveness level. The grid's carrying capacity attribute records the maximum number of tourists the grid can carry and the tourist movement speed corresponding to different tourist carrying capacities. The grid's reachability attribute records whether the grid can carry tourists.

3. The agent-based crowd flow simulation model according to claim 1, characterized in that, The basic attributes of tourists include their age, location, entry time, and exit time. The tourist's tour attributes include their energy and interest attributes. The energy attribute includes the tourist's energy level while visiting the scenic area. The interest attribute includes the degree of interest in attractions, the degree of interest decay, the degree of interest recovery, the minimum interest value, and the maximum interest value. The tourist's status attributes include information on attractions not visited, attractions visited, whether the tourist has moved to the target location, and whether the tourist has left the scenic area.

4. A method for simulating pedestrian flow, based on an agent-based pedestrian flow simulation model according to any one of claims 1-3, characterized in that, This includes initializing the entity model of the tourist and the regional model of the scenic area, predicting whether the tourist will continue to visit based on the exit decision; if not, determining the best exit, generating the navigation path of the scenic area and moving there; if yes, determining whether the tourist has entered the grid for the first time. Upon entering the grid area for the first time, the system predicts the attractions tourists will choose to visit based on attraction selection decisions, generates a navigation path, and moves accordingly. If this is not the first time entering the grid area, determine whether the tourist has reached the target attraction. If not, move along the existing navigation path. If yes, determine whether the tourist should stay and visit based on the attraction's dwell time decision. If a tourist stays to visit, a navigation path is generated based on the next location within the current attraction, and the tourist moves accordingly; if a tourist does not stay to visit, a navigation path is generated based on the attraction selection decision, predicting the attraction the tourist will choose to visit. Update visitor status and capabilities, then proceed to the step of predicting whether the visitor will continue their tour based on exit decisions.

Images