Pedestrian flow modeling and real-time simulation method, system, equipment and medium

Abstract

The invention relates to a pedestrian flow modeling and real-time simulation method, system and device and a medium, and the method comprises the following steps: carrying out the probability analysis of the moving direction of pedestrians in a railway platform environment based on a cellular automaton theory, and carrying out the modeling; based on the probability model of the pedestrian moving direction, according to the pedestrian decision-making process, modeling the pedestrian decision-making process by adopting a modeling method based on a multi-agent system; and simulating the pedestrian decision model by using an iterative vector quantization Q-Learning algorithm to obtain the overall distribution condition of the pedestrian flow for the next intelligent analysis of the digital twin system. The method can be widely applied to the computer and automation field and the railway video monitoring field.

Description

Technical Field

[0001] The present invention relates to a pedestrian flow modeling and real-time simulation method, system, device and medium, and in particular to a pedestrian flow modeling and real-time simulation method, system, device and medium of a multi-agent system (MAS) in a digital twin, and belongs to the fields of computer and automation, and railway video surveillance. Background Art

[0002] At present, railway video surveillance systems are widely deployed in railway platforms and other key areas with dense crowds and large passenger flows. They are an important measure to ensure passenger safety. They can effectively monitor and manage crowds, prevent potential risks, and ensure public safety.

[0003] Digital twin technology can add a layer of intelligent protection to this safety system. It creates a virtual copy of the actual scene through high-precision modeling, allowing us to monitor and analyze potential dangers in the real environment in real time in a simulated environment. This technology allows us to accurately map the actual scene in a multi-dimensional virtual space, and through this mapping, we can achieve real-time hazard warnings and optimize emergency escape route decisions. All of this is based on an accurate digital twin model, which is not only the core of functional realization, but also the basis for building an interactive environment.

[0004] The effectiveness of digital twin models depends on whether they can accurately reflect the dynamic changes of the real world in real time. This requires digital twin models to be constantly updated and iterated to keep pace with the physical world. By integrating advanced sensor data, real-time video analysis and artificial intelligence algorithms, digital twin models can provide dynamic scene analysis, predict and respond to various situations in real time, thereby greatly improving the management efficiency and response speed of public safety.

[0005] At present, digital twin technology is mainly used in modeling inanimate objects, such as scene reproduction, industrial machinery and equipment, and vehicle behavior simulation. In these systems, the complexity of work scheduling and process management is relatively simplified. However, modeling pedestrian flow involves many challenges: 1. Diversity of individual behavior: Each person’s behavior is affected by many factors, such as personal preferences, physical conditions, and psychological state, and individual behavior needs to be independently predicted; 2. Environmental factors: Pedestrian flow will be affected by environmental factors, including spatial layout, time, weather conditions, etc.; 3. Adaptability to dynamic changes: Pedestrian flow is dynamically changing, and the model needs to be able to adapt to or predict these changes in real time. Summary of the invention

[0006] In response to the above problems, the purpose of the present invention is to provide a pedestrian flow modeling and real-time simulation method, system, equipment and medium, design a pedestrian flow management simulation model for the digital twin environment, and combine the multi-agent system (MAS) and reinforcement learning technology to simulate the pedestrian movement scenes in the video surveillance system.

[0007] To achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a pedestrian flow modeling and real-time simulation method, comprising the following steps: Based on the cellular automaton theory, the pedestrian movement direction in the railway platform environment is analyzed and modeled probabilistically; Based on the probability model of pedestrian movement direction, the pedestrian decision-making process is modeled by adopting the modeling method based on multi-agent system; The pedestrian decision-making model is simulated using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow, which is used for the next step of intelligent analysis of the digital twin system.

[0008] Furthermore, the probabilistic analysis and modeling of the pedestrian movement direction in the railway platform environment based on the cellular automation theory includes: The railway platform environment is discretized based on cellular automata theory and simulated with a regularized network. Construct a cellular automaton model for simulating pedestrian movement behavior, and represent the pedestrian movement direction with probability based on preset movement rules; Based on the probabilistic representation of the pedestrian's moving direction, a pre-built probability model is used to determine the expected utility of the neighboring cells during the pedestrian's walking process.

[0009] Furthermore, the probability representation based on the pedestrian's moving direction uses a pre-built probability model to determine the expected utility of the neighboring cells during the pedestrian's walking process, including: The walking process of pedestrians is analyzed, and the expected utility of neighboring cells during the walking process is divided into movement utility and comfort utility. A probability model is constructed based on the Bayesian Nash equilibrium theory to describe the probability distribution of pedestrians' decisions about other pedestrians around them; Based on the probability of the pedestrian's moving direction and the probability model, the expected utility of each neighboring cell during the pedestrian's walking process is calculated.

[0010] Further, the expected utility is expressed as:

[0011] in, Indicates the utility of the pedestrian from the target point, which is the moving distance and moving direction Function of represents the comfort utility, and n represents the crowd density.

[0012] Further, the probability model is expressed as:

[0013] in, , , The population density , , The probability of a pedestrian entering a cell in these three cases; Due to the population density A constant that changes with the influence of.

[0014] Furthermore, the probability model based on the pedestrian's moving direction is based on the pedestrian's decision-making process and adopts a modeling method based on a multi-agent system to model the pedestrian's decision-making process, including: ① Before each decision cycle begins, obtain key parameters as input; ② Calculate the expected utility values ​​of the six neighboring cells around the pedestrian based on the key parameters obtained and factors such as the pedestrian’s current location and planned direction of travel; ③ Take the cell corresponding to the highest expected utility value as the target cell, and determine whether moving to the target cell will lead to too high a flow density or obstruction of movement. If so, randomly choose to move to other cells or stay in place until the next decision cycle; ④ Repeat steps ① to ③ until the pedestrian successfully reaches his or her predetermined target location and completes his or her journey.

[0015] Furthermore, when the pedestrian decision model is simulated using the iterative vector quantization Q-Learning algorithm, the reward function Q is defined as:

[0016] in, , Indicated in The states and actions of the multi-agent model are always observable. represents the learning rate, Represents the importance of reward, which is a constant less than 1.

[0017] In a second aspect, the present invention provides a pedestrian flow modeling and real-time simulation system, comprising: Probabilistic analysis and modeling module, which is used to perform probabilistic analysis and modeling of pedestrian movement directions in railway platform environments based on cellular automaton theory; The decision analysis and modeling module is used to model the pedestrian decision process based on the probability model of the pedestrian's movement direction and the modeling method based on the multi-agent system; The simulation module is used to simulate the pedestrian decision-making model using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow for the next step of intelligent analysis of the digital twin system.

[0018] In a third aspect, the present invention provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform any method.

[0019] In a fourth aspect, the present invention provides a computing device, comprising: one or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, and the one or more programs include instructions for executing any method.

[0020] The present invention adopts the above technical solution, which has the following advantages: 1. This invention is based on a multi-agent system and combines it with a reinforcement learning algorithm. Through this method, we can train a multi-agent model to accurately and in real time reflect the dynamics of the physical world. This advancement addresses the limitations of current technology in the use of digital twin data, especially in terms of fit with the real world and lack of real-time capabilities.

[0021] 2. Through data intercommunication with hardware facilities such as video surveillance and sensors in the physical world, the present invention can provide key information for the digital twin system to support decision-making, early warning response and other functions.

[0022] In summary, the present invention significantly improves the intelligence level of the video surveillance system and opens up new application prospects for the development of digital twin technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings: Figure 1 is a schematic diagram of the probability of pedestrians selecting adjacent cells provided in an embodiment of the present invention; Figure 2 It is a pedestrian flow modeling framework based on a multi-agent system provided in an embodiment of the present invention; Figure 3 It is a pedestrian decision flow chart provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0025] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0026] In some embodiments of the present invention, a pedestrian flow modeling and real-time simulation method is provided, which designs a pedestrian flow management simulation model for a digital twin environment, and combines a multi-agent system and reinforcement learning technology to simulate the pedestrian movement scene in a video surveillance system. In the pedestrian flow management simulation model, a multi-agent system is used to simulate individual pedestrians, and each agent can make independent decisions based on a series of complex variables to simulate the behavior of pedestrians. These variables include environmental factors (such as weather conditions), dynamic factors (such as train schedules and passenger flow), and psychological factors of pedestrians (such as personal sensitivity to crowding). Such multi-dimensional considerations enable the pedestrian flow management simulation model to more accurately predict the dynamic changes of pedestrian flow in different scenarios. The deep machine learning mechanism enables the model to be self-improving, and the accuracy and efficiency of the model are continuously improved with the accumulation of time and data to meet the accuracy and real-time requirements of the digital twin system.

[0027] Correspondingly, in some other embodiments of the present invention, a pedestrian flow modeling and real-time simulation system, device and medium are provided.

[0028] Example 1 like Figure 1 to Figure 3 As shown, the present invention provides a pedestrian flow modeling and real-time simulation method, which includes the following steps: 1) Based on the cellular automaton theory, the pedestrian movement direction in the railway platform environment is probabilistically analyzed and modeled; 2) Based on the probability model of pedestrian movement direction and the pedestrian decision-making process, a modeling method based on multi-agent system is used to model the pedestrian decision-making process; 3) The pedestrian decision-making model is simulated using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow for the next step of intelligent analysis of the digital twin system.

[0029] Furthermore, in the above step 1), for the railway video surveillance system, the present invention adopts Multi-Agent System (MAS) technology and incorporates the concept of cellular automata to accurately model the movement direction of pedestrians in the railway platform environment. Under this framework, pedestrians in the complex system are abstracted as agents with autonomous decision-making functions, which can move according to a predefined set of movement rules (including environmental feedback and interaction with other agents).

[0030] Cellular automata theory constitutes a discrete computational framework that simulates complex systems by building a regularized network. In the network, each unit cell is defined as a cell, which evolves iteratively according to preset rules. In this computational framework, time is regarded as a discrete quantity, and all cells follow a consistent update rule. The cells in the network will synchronously update their states in each discrete time step.

[0031] Specifically, the steps include: 1.1) Discretize the railway platform environment based on cellular automaton theory and simulate the railway platform environment with a regularized network; 1.2) Construct a cellular automaton model for simulating pedestrian motion behavior and represent the pedestrian's movement direction with probability based on preset movement rules; 1.3) Based on the probabilistic representation of the pedestrian’s moving direction, a pre-built probability model is used to determine the expected utility of the neighboring cells during the pedestrian’s walking process.

[0032] Furthermore, in the above step 1.2), in the pedestrian flow modeling problem of the multi-agent system, a movement rule needs to be formulated. This movement rule involves calculating the probability of pedestrians moving in different directions under an ideal environment and applying these probabilities to the update process of the cell state.

[0033] like Figure 1 As shown in the figure, the present invention constructs a cellular automaton model to simulate the movement behavior of pedestrians. Based on previous research on pedestrian density, the present invention assumes that the side length of each cell is 0.5m. In view of the fact that the probability of pedestrians moving backward is relatively low in actual situations, the cellular automaton model assumes that pedestrians move in five possible directions around in the next time step. Specifically, the probability of pedestrians moving to each direction is p 1. p 2. p 3. p 4. p5. In addition, the probability that the pedestrian remains at the current position is set to p 0. This probability setting reflects the realistic logic of pedestrian movement tendency and is parameterized accordingly in the cellular automaton model.

[0034] Furthermore, the above step 1.3) includes the following steps: 1.3.1) The walking process of pedestrians is analyzed, and the expected utility of the neighboring cells during the walking process is divided into movement utility and comfort utility.

[0035] In this embodiment, the expected utility of pedestrians during walking is divided into moving utility and comfort utility , the behavior of pedestrians during walking is the process of maximizing the total utility goal. Among them, the function expression of total utility is:

[0036] in, The utility of the pedestrian from the target point can be defined as the moving distance and moving direction The function is:

[0037] According to previous research on the impact of crowd crowding on comfort, It should contain a constant that varies with the crowd density. Let this constant be . Assume that the crowd density is n people / m 2 , It can be expressed as follows:

[0038] 1.3.2) A probability model is constructed based on the Bayesian Nash equilibrium theory to describe the probability distribution of pedestrians’ decisions about other pedestrians around them.

[0039] When analyzing pedestrians' walking behavior, we can use concepts from game theory to understand their decision-making process. Specifically, each pedestrian's choice is autonomous to some extent, that is, they are not fully aware of the intentions and strategies of other pedestrians around them when making movement decisions. In this case, pedestrians' decisions can be regarded as being made in an environment with incomplete information. In this context, Bayesian Nash equilibrium provides an analytical framework that allows us to predict the optimal behavior of pedestrians under uncertainty.

[0040] In the framework of Bayesian Nash equilibrium analysis, the present invention proposes to form one's own beliefs based on the prior probability distribution of other pedestrians' decisions around, and calculate the expected utility based on these beliefs. Subsequently, pedestrians will explore their own strategy space to seek a strategy that maximizes the expected utility. In the problem of pedestrian walking, the purpose of the present invention is to identify a specific cell through which pedestrians can achieve optimal path planning so as to efficiently reach the predetermined target location.

[0041] Therefore, the present invention constructs a probability model to describe the probability distribution of pedestrians' decisions about other pedestrians around them, and a method for calculating the expected utility based on this.

[0042] According to the Bayesian Nash equilibrium theory, considering the interaction between crowd density and comfort utility described in equation (3), the present invention defines the pedestrian comfort utility function as follows:

[0043] in, , , The population density , , The probability of a pedestrian entering a cell in these three cases.

[0044] 1.3.3) Based on the probability of the pedestrian’s moving direction and the probability model, the expected utility of each neighboring cell during the pedestrian’s walking process is calculated.

[0045] Since the probabilities of pedestrians entering a cell are p 0. p 1. p 2. p 3. p 4. p 5. Assume that there are six groups of people participating in the decision-making pedestrian game, and the number of pedestrians in each group is g 0. g 1. g 2. g 3. g 4. g 5. Since the cell size is set to 0.5m, it can be seen that the number of pedestrians in the same cell cannot be greater than 2, so and According to the probability distribution, the crowd density The probability is:

[0046] Crowd density When , there need to be two pedestrians around to choose the same cell. According to the probability distribution, we can get:

[0047] For the case where the crowd density is between 4 and 7, the dynamic characteristics are more complicated. In order to simplify the analysis, the present invention first focuses on the scene where the crowd density is equal to 9. At this density level, the crowd movement is significantly restricted, causing the pedestrian speed to drop to zero, forming a stagnant state. For the case of a density of 9, its implementation condition requires that pedestrians in the surrounding 5 or 6 cells must unanimously choose the same cell to gather. Based on this, it can be deduced that the probability of pedestrians choosing the same cell in this scenario is:

[0048] Furthermore, in the above step 2), if Figure 2 and Figure 3 As shown, the following steps are included: 2.1) Before each decision cycle begins, key parameters are obtained as input.

[0049] In this embodiment, the key parameters obtained include the size parameters of the space (such as length and width) and dynamic information such as crowd density.

[0050] 2.2) Based on the key parameters obtained and factors such as the pedestrian’s current location and planned direction of travel, the expected utility values ​​of the six neighboring cells around the pedestrian are calculated.

[0051] 2.3) The cell corresponding to the highest expected utility value is taken as the target cell, and whether it is possible to stay, that is, whether moving to the target cell will lead to too high a density of people or obstruction of travel. If so, randomly choose to move to other cells or stay in place until the next decision cycle; 2.4) Repeat steps 2.1) to 2.3) until the pedestrian successfully reaches his or her predetermined target location and completes his or her journey.

[0052] Furthermore, in the above step 3), for the decision model of each pedestrian agent, the present invention proposes an improved iterative vector quantization Q-Learning algorithm. The algorithm combines the advantages of the traditional Q-Learning framework and vector quantization technology, aiming to improve the decision-making quality and learning efficiency of the agent in a complex environment. Through continuous interaction and feedback with real-world data, the algorithm can continuously optimize its strategy performance, thereby more accurately simulating and predicting the behavior of pedestrians in specific environments.

[0053] In this algorithm, the reward function Q is defined as:

[0054] in, , Indicated in The states and actions of the multi-agent model are always observable. represents the learning rate, Indicates the importance of the reward, usually a constant less than 1.

[0055] In a complex dynamic environment, pedestrians, as entities making autonomous decisions, need to conduct multi-dimensional utility evaluations in each decision cycle to optimize their travel paths and speeds. Specifically, pedestrians will first use a pre-set utility function (as shown in Equation 4) to quantify their preferences, and then derive a vector representation of the maximum expected utility function. This process involves predicting and evaluating the possible results of different behavioral choices and converting them into computable state space vectors for subsequent optimization and selection. The detailed construction and algorithm flow of the decision model have been fully described in Table 1, and its logical structure and execution steps have been clearly divided.

[0056] Table 1 Pseudocode of multi-agent IT-VQQL algorithm

[0057] Example 2 The above-mentioned embodiment 1 provides a pedestrian flow modeling and real-time simulation method. Correspondingly, this embodiment provides a pedestrian flow modeling and real-time simulation system. The system provided in this embodiment can implement the pedestrian flow modeling and real-time simulation method of embodiment 1, and the system can be implemented by software, hardware, or a combination of software and hardware. For example, the system may include integrated or separate functional modules or functional units to execute the corresponding steps in each method of embodiment 1. Since the system of this embodiment is basically similar to the method embodiment, the process described in this embodiment is relatively simple. For relevant parts, please refer to the partial description of embodiment 1. The embodiment of the system provided in this embodiment is only schematic.

[0058] The pedestrian flow modeling and real-time simulation system provided in this embodiment includes: Probabilistic analysis and modeling module, which is used to perform probabilistic analysis and modeling of pedestrian movement directions in railway platform environments based on cellular automaton theory; The decision analysis and modeling module is used to model the pedestrian decision process based on the probability model of the pedestrian's movement direction and the modeling method based on the multi-agent system; The simulation module is used to simulate the pedestrian decision-making model using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow for the next step of intelligent analysis of the digital twin system.

[0059] Example 3 This embodiment provides a processing device corresponding to the pedestrian flow modeling and real-time simulation method provided in this embodiment 1. The processing device can be a processing device for a client, such as a mobile phone, a laptop computer, a tablet computer, a desktop computer, etc., to execute the method of embodiment 1.

[0060] The processing device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. The memory stores a computer program that can be run on the processor, and the processor executes the pedestrian flow modeling and real-time simulation method provided in this embodiment 1 when running the computer program.

[0061] Preferably, the memory may be a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory.

[0062] Preferably, the processor may be a central processing unit (CPU), a digital signal processor (DSP) or other general-purpose processors, which are not limited here.

[0063] Example 4 The pedestrian flow modeling and real-time simulation method of this embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the pedestrian flow modeling and real-time simulation method described in this embodiment 1 are loaded.

[0064] Computer readable storage media can be tangible devices that hold and store instructions used by instruction execution devices. Computer readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any combination thereof.

[0065] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 One or more processes and / or boxes Figure 1 A function specified in one or more boxes. These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A pedestrian flow modeling and real-time simulation method, characterized in that: The following steps are involved: Based on the cellular automaton theory, the pedestrian movement direction in the railway platform environment is analyzed and modeled probabilistically; Based on the probability model of pedestrian movement direction, the pedestrian decision-making process is modeled by adopting the modeling method based on multi-agent system; The pedestrian decision-making model is simulated using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow, which is used for the next step of intelligent analysis of the digital twin system.

2. A pedestrian flow modeling and real-time simulation method as claimed in claim 1, characterized in that: Based on the cellular automaton theory, the pedestrian movement direction in the railway platform environment is analyzed and modeled probabilistically, including: The railway platform environment is discretized based on cellular automata theory and simulated with a regularized network. Construct a cellular automaton model for simulating pedestrian movement behavior, and represent the pedestrian movement direction with probability based on preset movement rules; Based on the probabilistic representation of the pedestrian's moving direction, a pre-built probability model is used to determine the expected utility of the neighboring cells during the pedestrian's walking process.

3. A pedestrian flow modeling and real-time simulation method as claimed in claim 2, characterized in that: The probability representation based on the pedestrian's moving direction uses a pre-built probability model to determine the expected utility of the neighboring cells during the pedestrian's walking process, including: The walking process of pedestrians is analyzed, and the expected utility of neighboring cells during the walking process is divided into movement utility and comfort utility. A probability model is constructed based on the Bayesian Nash equilibrium theory to describe the probability distribution of pedestrians' decisions about other pedestrians around them; Based on the probability of the pedestrian's moving direction and the probability model, the expected utility of each neighboring cell during the pedestrian's walking process is calculated.

4. A pedestrian flow modeling and real-time simulation method as claimed in claim 3, characterized in that: The expected utility is expressed as: in, Indicates the utility of the pedestrian from the target point, which is the moving distance and moving direction Function of represents the comfort utility; n represents the crowd density.

5. The pedestrian flow modeling and real-time simulation method according to claim 3, characterized in that: The probability model is expressed as: in, , , The population density , , The probability of a pedestrian entering a cell in these three cases; is a constant that varies with the density of the crowd.

6. A pedestrian flow modeling and real-time simulation method as claimed in claim 1, characterized in that: The probability model based on the pedestrian's moving direction is based on the pedestrian's decision-making process and adopts a modeling method based on a multi-agent system to model the pedestrian's decision-making process, including: ① Before each decision cycle begins, obtain key parameters as input; ② Calculate the expected utility values ​​of the six neighboring cells around the pedestrian based on the key parameters obtained and factors such as the pedestrian’s current location and planned direction of travel; ③ Take the cell corresponding to the highest expected utility value as the target cell, and determine whether moving to the target cell will lead to too high a density of people or obstruction of movement. If so, randomly choose to move to other cells or stay in place until the next decision cycle; ④ Repeat steps ① to ③ until the pedestrian successfully reaches his or her predetermined target location and completes his or her journey.

7. The pedestrian flow modeling and real-time simulation method according to claim 1, characterized in that: When the pedestrian decision model is simulated using the iterative vector quantization Q-Learning algorithm, the reward function Q is defined as: in, , Indicated in The states and actions of the multi-agent model are always observable. represents the learning rate, Represents the importance of reward, which is a constant less than 1.

8. A pedestrian flow modeling and real-time simulation system, characterized in that: include: Probabilistic analysis and modeling module, which is used to perform probabilistic analysis and modeling of pedestrian movement directions in railway platform environments based on cellular automaton theory; The decision analysis and modeling module is used to model the pedestrian decision process based on the probability model of the pedestrian's movement direction and the modeling method based on the multi-agent system; The simulation module is used to simulate the pedestrian decision-making model using the iterative vector quantization Q-Learning algorithm to obtain the overall distribution of pedestrian flow for the next step of intelligent analysis of the digital twin system.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform any one of the methods of claims 1 to 7.

10. A computing device, characterized in that include: One or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, wherein the one or more programs include instructions for executing any one of the methods described in claims 1 to 7.

Images