A crowd evacuation method based on coupling of emotional contagion and social force model

Abstract

The application is suitable for the field of crowd simulation and emergency evacuation, and provides a crowd evacuation method based on emotion contagion and social force model coupling, which comprises the following steps: constructing an emotion contagion model based on SEIRS framework, dividing individual emotion state into susceptible state, exposed state, infected state and recovered state, and dynamically adjusting infection rate and recovery rate through feedback function; establishing a social force model, embedding emotion value into target driving force and individual repulsive force calculation, realizing real-time adjustment of emotion on movement behavior; fusing global path planning and ant colony pheromone mechanism, introducing density perception cost function and pheromone guiding force, realizing dynamic congestion avoidance; designing a bidirectional feedback mechanism, triggering local path reconstruction when congestion is detected. The application realizes the bidirectional coupling of emotion transmission and behavior decision, improves the reality and adaptability of evacuation simulation, and can provide a scientific basis for public place emergency management and evacuation strategy optimization.

Description

Technical Field

[0001] This invention belongs to the field of crowd simulation and emergency evacuation technology, and in particular relates to a crowd evacuation method based on the coupling of emotional contagion and social force model. Background Technology

[0002] Currently, crowd evacuation simulation has become an important research area in the field of public safety. Existing technologies commonly employ crowd evacuation models, primarily including physics-driven social force models and emotion-based models based on infectious disease dynamics.

[0003] Social force models are among the most widely used evacuation modeling methods. Their basic idea is to abstract individual movement behavior as the result of multiple forces acting together, including the driving force of individuals toward a goal, the repulsive force between individuals, and the constraint force of environmental obstacles. This model can effectively simulate avoidance behavior between individuals, congestion, and channel bottleneck effects, and is therefore widely used in evacuation simulations of subway stations, stadiums, shopping malls, and other similar locations. However, traditional social force models primarily focus on physical mechanical interactions, lacking a characterization of the psychological and emotional states of the population. In actual emergency evacuations, individuals' panic, anxiety, and other emotions significantly influence their speed, direction selection, and avoidance tendencies, and models based solely on physical forces struggle to reflect this complex behavioral characteristic.

[0004] Some studies have attempted to combine emotion propagation models with evacuation behavior models, but they generally suffer from the following shortcomings:

[0005] 1. The coupling method is too simple: Some studies only set a fixed velocity gain to reflect the influence of emotions on motion, and failed to establish a reasonable function mapping relationship, resulting in simulation results that lack universality and flexibility.

[0006] 2. Lack of two-way feedback mechanism: Existing technologies often only consider "emotions affecting behavior" without considering that crowded environments can amplify the rate of emotion transmission. Therefore, they cannot accurately reproduce the complex process of the interaction and mutual promotion of emotions and behaviors in reality.

[0007] 3. Lack of dynamism in route selection: Many studies have shown that individual route planning is static and does not adjust according to local congestion and traffic conditions. Therefore, it is difficult to realistically reproduce the phenomenon of people changing their exit choices due to panic in emergency situations.

[0008] The root cause of these problems lies in the fact that traditional methods lack an effective two-way coupling framework between emotion and behavior, fail to quantify emotional states and embed them into the parameter system of social force models, and also lack a dynamic regulation mechanism that considers the impact of crowded environments on the rate of emotion transmission. Therefore, existing technologies cannot fully reflect the complex interaction between emotions and behaviors in emergency situations. Summary of the Invention

[0009] The purpose of this invention is to provide a crowd evacuation method based on the coupling of emotional contagion and social force models, in order to solve the problems mentioned in the background art.

[0010] The present invention is implemented as follows: a crowd evacuation method based on the coupling of emotional contagion and social force models includes the following steps:

[0011] Step 1: Construct an emotion contagion model, classifying individual emotional states into susceptible, exposed, infected, and recovered states. Based on the SEIRS infectious disease dynamics framework, establish a dynamically adjustable emotion state transition equation, where the infection rate... and recovery rate Through feedback function Real-time adjustments are made, and the feedback function is based on the individual's local crowding level. Duration of obstruction Distance from the group Calculated;

[0012] Step 2: Establish a social force model, assigning the individual's current emotional value. Embedded in the social force equation, dynamically adjust the goal-driving force and the repulsive force between individuals;

[0013] Step 3: Couple emotions with behaviors by using the individual's current emotional value. Adjust the expected speed and perceived distance in the social force model so that the individual's motor behavior adapts to changes in emotional state;

[0014] Step 4: Perform dynamic path planning and integration The pheromone mechanism of the algorithm and the ant colony algorithm (ACO) introduces a density-aware term into the path cost function and adds a guiding force based on pheromone gradient into the social force model;

[0015] Step 5: Introduce a two-way feedback mechanism. When the local congestion or pheromone concentration exceeds a preset threshold, trigger local path reconstruction to achieve dynamic congestion avoidance.

[0016] In a further technical solution, step 1 includes the following specific steps:

[0017] An infectious disease dynamics model was used to simulate the spread of panic in a population. Based on the traditional SEIRS model, a dynamically adjustable emotion transfer model was established to simulate the temporal evolution of panic among individuals, taking into account the characteristics of emotion transmission in a group evacuation scenario. Individual emotional states were categorized into four types: susceptible, exposed, infected, and recovered. Considering the spontaneous emotional contagion phenomenon in emergency events, the following state transition equation was constructed:

[0018] ;

[0019] in, Indicates the infection rate, reflecting the intensity of emotional transmission; The incubation period conversion rate; Recovery rate, reflecting the rate at which an individual recovers from a state of panic; For the rate of immune regression; Spontaneous infection rate, used to describe panic in individuals caused by sudden stimuli in non-contact settings; The proportion of individuals in a susceptible state; The proportion of individuals in the exposed state; The proportion of individuals in an infected state; The proportion of individuals in the recovery state; among which, the infection rate With recovery rate All follow the feedback function Dynamic adjustment It is calculated in real time from local congestion, the degree of speed obstruction, and the average group sentiment.

[0020] A motion feedback regulation mechanism is constructed to achieve a closed loop coupling between emotion transmission and behavioral decision-making; specifically, this is achieved by monitoring the local crowding level of individuals. Duration of obstruction Distance from the group Dynamically adjust the parameters for emotional state transfer so that the movement state can, in turn, influence the emotional transmission process.

[0021] Among them, local congestion By statistical individuals Surround perception radius Number of neighbors Divide by the area of ​​the region Obtain, that is Duration of obstruction The threshold is obtained by detecting whether the individual's velocity is below a certain threshold in consecutive frames during simulation. If the distance is lower, the time will be accumulated; group distance The calculation formula is ,in, For individuals Location coordinates, For individuals Neighbor location Position coordinates;

[0022] Introducing a feedback function This can lead to heightened emotions in individuals under conditions of high density and crowd congestion, further influencing their subsequent motor behavior; feedback function The expression is:

[0023] ;

[0024] in, To maximize local congestion, The maximum blocking time, For the maximum group distance, , , These are the weighting coefficients for the influence of three factors—local crowding, time of obstruction, and group distance—on individual emotions. These weighting coefficients satisfy the following: ;

[0025] Due to the transmission of emotions, the infection rate and recovery rate Due to local congestion Duration of obstruction Distance from the group The emotional state changes due to influences, and the formula for adjusting emotional state transition is:

[0026] ;

[0027] ;

[0028] in, The baseline infection rate for emotional transmission represents the basic transmission intensity under conditions without external stress. The baseline recovery rate represents the natural rate of calming down in a stress-free environment. This formula illustrates that as local congestion and duration of obstruction increase, the infection rate increases while the recovery rate decreases.

[0029] In a further technical solution, step 2 includes the following specific steps:

[0030] After clarifying the rules governing the contagion of panic, a social force model is introduced to simulate individuals' path planning and obstacle avoidance behaviors. The individual's emotional value, as their state output, is determined by their state within the emotion propagation model, further modulating the goal-driven force and inter-individual repulsive force in the social force model. The individual's current emotional value... Embedded as a continuous quantity in the mechanical equations, it dynamically influences the escape plan and spatial tolerance; based on emotion propagation, it incorporates the individual's current emotional value. The mapping is set to continuous values, serving as a moderating factor for multiple force terms in the social force model; the mapping is set according to the state of emotion propagation.

[0031] ;

[0032] in, This represents the individual's current emotional state. The initial emotion value in the recovery state. is the base of the exponential function. This represents the cumulative recovery time after entering the recovery state; the susceptible state indicates that the individual's emotions have not yet been infected, with the lowest emotion value, set to 0; the infected state indicates that the individual is completely in a state of panic, with an emotion value of 1; the exposed state indicates that the individual has been affected but is not completely panicked, with an emotion value between 0 and 1, using the median value of 0.5 here; the recovery state indicates that the individual has gradually recovered from a state of panic to calmness, and its emotion value is controlled by a decay function. This represents the initial emotional value of an individual when entering a recovery state; Let the emotional recovery constant be defined as follows: ,in It is the half-life of emotion.

[0033] A further technical solution is provided in step 2. The value is 0.7, and in public place crowd evacuation scenarios, Take 8 seconds, corresponding to .

[0034] In a further technical solution, step 3 includes the following specific steps:

[0035] The model improves upon this by embedding emotion values ​​into a social force model, in which individuals attempt to move at a desired speed. The formula for the driving force towards a goal is:

[0036] ;

[0037] in, It refers to individual quality; Individual Expected speed, It is the unit vector of the target direction. At the current speed, This is a measure of the time required for an individual to adjust its speed to a desired state. Driven by goals. Incorporating the individual's current emotional state. Afterwards, expected speed Adjusted to:

[0038] ;

[0039] in, The expected speed in a calm state, Let be the coefficient of influence of emotion value on speed; after substituting, the driving force of the goal becomes:

[0040] ;

[0041] In the social force model, there is a repulsive force between people, expressed by the formula:

[0042] ;

[0043] in, The repulsive force intensity constant is To perceive distance scales. For a safe distance, This is the actual distance. Indicates from point to unit vector, The repulsive force between people;

[0044] To reflect the impact of emotions on an individual's perceptual range and reaction intensity, the individual's current emotion value is introduced. The above formula is adjusted as follows:

[0045] ;

[0046] in, The emotion regulation coefficient is used to control the influence of emotion value on perceived distance. Through the above adjustments, the model can achieve adaptive control of individual rejection behavior as emotion state dynamically changes; thus, the general formula of the social force model is obtained:

[0047] ;

[0048] in, Represents an individual With the wall The forces between them For individuals The speed.

[0049] In a further technical solution, step 4 includes the following specific steps:

[0050] Based on the bidirectional coupling of the social force model and the emotion transmission model, the integration Path planning and ACO form a multi-agent evacuation simulation framework;

[0051] The pheromone mechanism from ACO is introduced and combined with local population density estimation to simulate the self-organized regulation of group path selection. Individuals continuously release pheromones into the current grid during movement, and the system updates node pheromone values ​​according to the following rules:

[0052] ;

[0053] in, For nodes exist The pheromone concentration at any given time; The pheromone evaporation coefficient is used to control its decay rate over time. For the first Individuals at nodes The increase in pheromones left on the surface;

[0054] Based on the static structural information of the evacuation environment, through classic... The algorithm generates a shortest path from the current location to the target exit for each individual; this path consists of several key navigation nodes, serving as the individual's initial target sequence and providing global navigation guidance for the system; during path planning, to avoid congested areas, [the algorithm further optimizes the path]. The algorithm's cost function has been improved by introducing density-aware weight terms:

[0055] ;

[0056] in, For the current node The total cost, From the starting point to the current node The cumulative actual cost, From Heuristic distance to the exit, This indicates the population density at the current node. This is the density influence factor, used to adjust the weight of local congestion in path cost;

[0057] Based on this, a guiding force based on pheromones is added to the social force model:

[0058] ;

[0059] in, For pheromone guidance, The gradient of the pheromone field points in the direction of the fastest increase in concentration; the social force model equation further becomes:

[0060] .

[0061] A further technical solution involves a dynamic path update mechanism in step 5: when the density or pheromone concentration in a certain area exceeds a preset threshold, the individual triggers local path reconstruction. The algorithm recalculates the local shortest path from the current position to the exit. The new path also considers density or pheromone cost, thereby avoiding the impact of local congestion or sudden environmental changes.

[0062] The crowd evacuation method based on the coupling of emotional contagion and social force model provided in this invention has the following beneficial effects:

[0063] (1) Two-way coupling of emotion and behavior, making the simulation closer to reality: By deeply integrating the emotion propagation model with the social force model, the real-time influence of emotional state on movement behavior and the reverse regulation of movement state on emotion propagation are realized, which significantly improves the psychological realism and behavioral rationality of crowd evacuation simulation.

[0064] (2) Dynamic adaptive path planning effectively alleviates congestion: combined with Global path planning and ant colony pheromone mechanism are introduced, along with density-aware cost function and pheromone guidance force, enabling individuals to perceive and avoid high-density areas in real time, improving evacuation efficiency and reducing congestion and deadlock.

[0065] (3) Refined modeling of emotion transmission mechanism to support multi-state evolution: Based on the SEIRS framework, an emotion transmission model is constructed, introducing the exposure state and the recovery state to simulate the process of the incubation, outbreak, dissipation and reinfection of panic, which is more in line with the complex evolution of people's emotions in actual emergency scenarios.

[0066] (4) The parameters are adjustable and highly scalable, and applicable to a variety of scenarios: Key parameters in the model (such as emotional half-life, influence coefficient, weight factor, etc.) can be calibrated and adjusted according to specific scenarios, and are applicable to simulation and optimization of various densely populated environments such as public places, transportation hubs, and emergency drills.

[0067] (5) The system responds quickly and supports dynamic environment adaptation: Through the bidirectional feedback mechanism and local path reconstruction capability, the system can quickly respond to environmental changes (such as sudden congestion, exit closure, etc.), and improve the dynamic adaptability and robustness of the overall simulation.

[0068] (6) Provide scientific basis for emergency plans and facility design: This method can be used to evaluate the effectiveness of different evacuation strategies, exit layouts and guidance measures, and provide data support and decision-making reference for the safety design and emergency management of public places. Attached Figure Description

[0069] Figure 1 A flowchart illustrating the two-way coupling of emotion propagation and behavioral decision-making;

[0070] Figure 2 To integrate population density Path planning optimization algorithm flow;

[0071] Figure 3 This is the emotional contagion process of the state transition equation system. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0073] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0074] like Figure 1 and Figure 2 As shown, an embodiment of the present invention provides a crowd evacuation method based on the coupling of emotional contagion and social force models, comprising the following steps:

[0075] Step 1: Construct an emotion contagion model;

[0076] To better illustrate the spread of panic within a population, an infectious disease dynamics model is employed. Since emotion transmission within a group also exhibits a dynamic process of "infection—recovery—reinfection," a dynamically adjustable emotion transfer model is established based on the traditional SEIRS model, specifically tailored to the characteristics of emotion transmission in group evacuation scenarios. This model characterizes the temporal evolution of panic among individuals. Individual emotional states are categorized into four types: susceptible, exposed, infected, and recovered. Considering the spontaneous emotional contagion phenomenon present in emergency events, the following state transition equation is constructed:

[0077] ;

[0078] in, Indicates the infection rate, reflecting the intensity of emotional transmission; The incubation period conversion rate; Recovery rate, reflecting the rate at which an individual recovers from a state of panic; For the rate of immune regression; Spontaneous infection rate, used to describe panic in individuals caused by sudden stimuli in non-contact settings; The proportion of individuals in a susceptible state; The proportion of individuals in the exposed state; The proportion of individuals in an infected state; This represents the proportion of individuals in the recovery state. Among them, the infection rate... With recovery rate All can follow the feedback function Dynamic adjustment It can be calculated in real time from factors such as local congestion, the degree of speed obstruction, and the average group sentiment.

[0079] Since not only can the spread of panic influence individual movement decisions, but an individual's movement state can also, in turn, affect the spread process, a movement feedback regulation mechanism is constructed to achieve a closed loop coupling between emotion transmission and behavioral decision-making. Specifically, this is achieved by monitoring the local crowding level of individuals. Duration of obstruction Distance from the group By dynamically adjusting parameters for emotional state transfer based on key behavioral indicators, the movement state can, in turn, influence the emotional transmission process.

[0080] Among them, local congestion By statistical individuals Surround perception radius Number of neighbors Divide by the area of ​​the region Obtain, that is Duration of obstruction The threshold is obtained by detecting whether the individual's velocity is below a certain threshold in consecutive frames during simulation. If the distance is lower, the time will be accumulated; group distance The calculation formula is ,in, For individuals Location coordinates, For individuals Neighbor location The location coordinates.

[0081] Introducing a feedback function This can lead to heightened emotions in individuals under conditions of high density and crowd congestion, further influencing their subsequent motor behavior. Feedback function The expression is:

[0082] ;

[0083] in, To maximize local congestion, The maximum blocking time, For the maximum group distance, , , These are the weighting coefficients for the influence of three factors—local crowding, time of obstruction, and group distance—on individual emotions. These weighting coefficients satisfy the following: .

[0084] Due to the transmission of emotions, the infection rate and recovery rate Due to local congestion Duration of obstruction Distance from the group The emotional state changes due to influences, and the formula for adjusting emotional state transition is:

[0085] ;

[0086] ;

[0087] in, The baseline infection rate for emotional transmission represents the basic transmission intensity under conditions without external stress. The baseline recovery rate represents the natural rate of calming down in a stress-free environment; this formula illustrates that as local congestion and duration of obstruction increase, the infection rate increases and the recovery rate decreases.

[0088] The emotional contagion process embodied in this set of state transition equations is as follows: Figure 3 As shown in the diagram, the process can be categorized as follows: "Susceptible state" → "Exposed state" represents an individual experiencing potential anxiety but not yet completely out of control due to the influence of surrounding sources of panic; "Exposed state" → "Infected state" represents complete panic infection after an incubation period; "Infected state" → "Recovery state" represents the gradual subsiding of panic into calm; and "Recovery state" → "Susceptible state" represents the process of a calm individual returning to susceptibility or being reinfected without external stimuli. The introduction of the "Exposed state" simulates the incubation period of panic, while the inclusion of the "Recovery state" and "Recovery state" → "Susceptible state" pathways reflects the reinfection mechanism.

[0089] Step 2: Establish a social force model;

[0090] After clarifying the rules governing the contagion of panic, a social force model is introduced to simulate individuals' path planning and obstacle avoidance behaviors. The individual's emotional value, as their state output, is determined by their state within the emotion propagation model, further modulating the goal-driven force and inter-individual repulsive force within the social force model. The individual's current emotional value... Embedded as a continuous quantity in the mechanical equations, it dynamically influences the escape plan and spatial tolerance. Based on emotion propagation, it incorporates the individual's current emotional value. The mapping is done as continuous values, serving as a moderating factor for multiple force terms in the social force model. The mapping is set according to the state of emotion propagation:

[0091] ;

[0092] in, This represents the individual's current emotional state. The initial emotion value in the recovery state. is the base of the exponential function. This represents the cumulative recovery time after entering the recovery state; the susceptible state indicates that the individual's emotions have not yet been infected, with the lowest emotion value, set to 0; the infected state indicates that the individual is completely in a state of panic, with an emotion value of 1; the exposed state indicates that the individual has been affected but is not completely panicked, with an emotion value between 0 and 1, here using the median value of 0.5; the recovery state indicates that the individual has gradually recovered from a state of panic to calmness, and its emotion value is controlled by a decay function. This represents the initial emotional value of an individual when entering a recovery state, due to the individual's initial emotional state. (1.0) Back to It won't immediately return to the way it was. Such a low level, which is set to 0.7 in this formula, is used to characterize that although the individual has escaped the state of panic, he still maintains a certain level of tension. Let the emotional recovery constant be defined as follows: ,in The half-life of emotions occurs during crowd evacuation in public places. Take 8 seconds, corresponding to However, this invention is not limited to these values; the parameters can be calibrated according to the scenario and data. The above design can achieve exponential decay of emotional values ​​during the recovery process, reflecting the gradual stabilization of psychological changes after an individual rapidly relaxes under high-pressure situations, thereby improving the realism and psychological rationality of emotion transmission modeling.

[0093] Step 3: Emotion-behavior coupling;

[0094] This is improved by embedding emotion values ​​into the social force model. In the social force model, individuals attempt to move at a desired speed. The formula for the driving force towards a goal is:

[0095] ;

[0096] in, It refers to individual quality; Individual Expected speed, It is the unit vector of the target direction. At the current speed, This is a measure of the time required for an individual to adjust its speed to a desired state. Driven by goals. Incorporating the individual's current emotional state. Afterwards, expected speed Adjusted to:

[0097] ;

[0098] in, The expected speed in a calm state, This is the coefficient representing the influence of emotional value on speed. This adjustment reflects the fact that the more panicked the emotion (…), the greater the impact on speed. The more an individual wants to escape faster, the higher their expected speed becomes. Substituting this into the equation, the goal-driven force becomes:

[0099] ;

[0100] In the social force model, there is a repulsive force between people, expressed by the formula:

[0101] ;

[0102] in, The repulsive force intensity constant is To perceive distance scales. The safe distance is the sum of the radii of the two people. This is the actual distance. Indicates from point to unit vector, It is the force of repulsion between people.

[0103] To reflect the impact of emotions on an individual's perceptual range and reaction intensity, the individual's current emotion value is introduced. The above formula is adjusted as follows:

[0104] ;

[0105] in, This is the emotion regulation coefficient, used to control for the impact of emotion value on perceived distance. When an individual's emotion value is high (i.e., in a state of panic), Increase Increasing the denominator and decreasing the exponent term results in a slower decay of the repulsive force, thus expanding its effective range and reflecting an individual's high sensitivity to approach from others. When an individual's emotional state is low, this term approaches 1, reverting to the traditional social force model. Through these adjustments, the model can achieve adaptive control of individual rejection behavior dynamically changing with emotional state, making crowd evacuation simulations more closely resemble the psychological and behavioral responses in actual emergency scenarios. This leads to the general formula for the social force model:

[0106] ;

[0107] in, Represents an individual With the wall The forces between them For individuals The speed.

[0108] Step 4: Dynamic path planning;

[0109] To address the limitations of traditional path planning in high-density, dynamic crowd dispersal scenarios, which are prone to congestion and lack adaptability, this paper integrates a social force model and an emotion propagation model, based on the bidirectional coupling of these two models. Path planning and ACO form a multi-agent evacuation simulation framework.

[0110] To further enhance the adaptability and group coordination of path selection, the pheromone mechanism from ACO was introduced and combined with local population density estimation to simulate the self-organizing regulation of group path selection. Individuals continuously release pheromones into the current grid during movement, and the system updates node pheromone values ​​according to the following rules:

[0111] ;

[0112] in, For nodes exist The pheromone concentration at any given time; The pheromone evaporation coefficient is used to control its decay rate over time. For the first Individuals at nodes The incremental pheromone left behind. Through the spatial distribution and gradient of the pheromone field, individuals can perceive the density of people and their travel experience when choosing a path, thus avoiding congestion.

[0113] Based on the static structural information of the evacuation environment, through classic... The algorithm generates a shortest path from the current location to the target exit for each individual. This path consists of several key navigation nodes, serving as the individual's initial target sequence and providing global navigation guidance for the system. During path planning, to avoid congested areas, [the algorithm...]. The algorithm's cost function has been improved by introducing density-aware weight terms:

[0114] ;

[0115] in, For the current node The total cost, From the starting point to the current node The cumulative actual cost, From Heuristic distance to the exit, This represents the population density (pheromone concentration) at the current node. This is a density influence factor used to adjust the weight of local congestion in path cost, considering the combined effect of distance and density. The impact of path planning is taken as a value here. This is to guide individuals to avoid high-density areas.

[0116] Based on this, the social force model can be further improved by adding a guiding force based on pheromones:

[0117] ;

[0118] in, For pheromone guidance, Let be the gradient of the pheromone field, pointing in the direction of the fastest increase in concentration. That is, during evacuation, people instinctively avoid areas with high pheromone concentrations and walk in the opposite direction of the pheromone gradient. Therefore, the social force model equation further becomes:

[0119] ;

[0120] Step 5: Introduce a two-way feedback mechanism;

[0121] A dynamic path update mechanism was designed: when the density or pheromone concentration in a certain area exceeds a preset threshold, the individual triggers local path reconstruction, through... The algorithm recalculates the local shortest path from the current position to the exit. The new path also considers density or pheromone costs, thus effectively avoiding the impact of local congestion or sudden environmental changes. This mechanism significantly improves the system's dynamic response capability and overall simulation stability in complex environments.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A crowd evacuation method based on coupling of emotional contagion and social force model, characterized in that, The method comprises the following steps: Step 1: Constructing the emotional contagion model, dividing the individual emotional state into susceptible state, exposed state, infected state and recovered state, and establishing the emotional state transfer equation which can dynamically adjust based on the SEIRS infectious disease dynamics framework, wherein the infection rate and the recovery rate are adjusted in real time through the feedback function based on the individual's local congestion , blocked time and group distance ; Step 2: Establishing the social force model, on the basis of emotion transmission, mapping the individual's current emotion value to a continuous value as the adjustment factor of multiple force terms in the social force model, embedding the individual's current emotion value in the social force equation to dynamically adjust the target driving force and repulsion between individuals; Step 3: Realize the coupling of emotion and behavior, through the individual's current emotion value Adjust the expected speed and perceived distance in the social force model, so that the individual's motion behavior changes adaptively with the emotional state; Step 4: Dynamic path planning is performed, fusing The density perception term is introduced in the path cost function and the guiding force based on pheromone gradient is added in the social force model by combining the algorithm and the pheromone mechanism of ACO. Step 5: A bidirectional feedback mechanism is introduced, when the local congestion degree or pheromone concentration exceeds a preset threshold, the individual local path reconstruction is triggered, and dynamic congestion avoidance is realized. 2.The crowd evacuation method based on the coupling of the emotional contagion and social force model according to claim 1, characterized in that, The step 1 comprises the following specific steps: An epidemic dynamics model is used to simulate the spread of panic in the crowd, and on the basis of the traditional SEIRS model, an emotion transfer model capable of dynamic adjustment is established for the emotion propagation characteristics in the crowd evacuation scene, so as to simulate the time sequence evolution process of panic emotion between individuals; wherein, the emotion state of the individual is divided into four categories: susceptible state, exposed state, infected state and recovered state, and the spontaneous emotion infection phenomenon existing in the emergency is considered, and the following state transition equation is constructed: ； in, Indicates the infection rate, reflecting the intensity of emotional transmission; The incubation period conversion rate; Recovery rate, reflecting the rate at which an individual recovers from a state of panic; For the rate of immune regression; Spontaneous infection rate, used to describe panic in individuals caused by sudden stimuli in non-contact settings; The proportion of individuals in a susceptible state; The proportion of individuals in the exposed state; The proportion of individuals in an infected state; The proportion of individuals in the recovery state; among which, the infection rate With recovery rate All follow the feedback function Dynamic adjustment It is calculated in real time from the weighted average of local congestion, obstruction time, and group distance. Constructing a motion feedback regulation mechanism to realize the coupling closed loop between emotion propagation and behavior decision; Specifically, by monitoring the local congestion degree of individuals , blocked time , and the distance from the group , dynamically adjust the emotion state transition parameters, so that the motion state in turn affects the emotion propagation process; where local congestion degree is obtained by counting the number of individuals within the perception radius of an individual , divided by the area of the region , i.e. ; the blocked time is obtained by detecting in the simulation whether the speed of an individual is lower than a threshold value in successive frames, and if so, the time is counted and accumulated; the group distance is calculated according to the formula , where is the position coordinate of an individual , and is the position coordinate of the neighbor position of the individual . Introducing feedback function , make the individual in high density and crowd congestion state emotional rise, further affect its subsequent sports behavior; feedback function The expression is: ； wherein, is the maximum local crowding, is the maximum blocked time, is the maximum group distance, , , are the weight coefficients of the three factors of local crowding, blocked time and group distance on individual emotion, respectively, and the weight coefficients satisfy: ; Since the infection rate and recovery rate will change due to local congestion , blocked time and group distance during the emotion propagation process, the emotion state transition formula is adjusted as follows: ； ； where, is the basic infection rate of emotion contagion, representing the basic transmission strength under no external stress condition; is the basic recovery rate of emotion recovery, representing the natural recovery rate of individuals in a stress-free environment; this equation shows that the infection rate increases and the recovery rate decreases with the increase of local congestion and blocked time. 3.The crowd evacuation method based on the coupling of the emotional contagion and social force model according to claim 2, characterized in that, The step 2 comprises the following specific steps: After the rules of panic emotion contagion are defined, the social force model is introduced to simulate the path planning and obstacle avoidance behavior of individuals; the emotion value as the state output of individuals is determined by its state in the emotion propagation model, and further adjusts the target driving force and repulsive force between individuals in the social force model; the current emotion value of individuals is embedded into the mechanical equation as a continuous quantity, dynamically affecting its escape intention and spatial tolerance; according to the emotion propagation state, the mapping is set: ； wherein, is the current emotional value of the individual, is the initial emotional value of the recovery state, is the base of the exponential function, is the cumulative recovery time after entering the recovery state; the susceptible state indicates that the emotion has not been infected, the emotional value is the lowest, and is set to 0; the infected state indicates that the individual is completely in a panic state, and the emotional value is 1; the exposed state indicates that the individual has been affected but not completely panicked, and the emotional value is between 0 and 1, and here the intermediate value 0.5 is adopted; the recovery state indicates that the individual gradually recovers from the panic state, and the emotional value is controlled by a decay function, represents the initial emotional value of the individual when entering the recovery state; is the emotional recovery constant, defined as wherein is the emotional half-life. 4.The crowd evacuation method based on the coupling of the emotional contagion and social force model according to claim 3, characterized in that, In the step 2, is 0.7, and in the public place crowd evacuation scenario, 8 seconds, corresponding to .

5. The crowd evacuation method based on coupling of emotional contagion and social force model according to claim 3, characterized in that, The step 3 comprises the following specific steps: The emotional value is embedded in the improved social force model, in which individuals try to move at the desired speed The formula of the target driving force is: ； wherein, is the individual mass; is the individual is the desired velocity, is the target direction unit vector, is the current velocity, denotes the amount of time measure the individual needs to adjust the velocity to the desired state, is the target driving force; the individual's current mood value is introduced after which the desired velocity is adjusted to: ； wherein, is the desired speed in a calm state, is the influence coefficient of the emotional value on the speed; after substitution, the target driving force becomes: ； In the social force model, there is a repulsive force between people, which is expressed by the formula: ； wherein, is a repulsion strength constant, is a perceived distance scale, is a safety distance, is an actual distance, denotes a unit vector from points to is a human-to-human repulsion force;​ In order to reflect the influence of emotion on the individual's perception range and response intensity, the current emotion value of the individual is introduced The above formula is adjusted as follows: ； wherein, is the emotion regulation coefficient, which is used to control the influence of the emotion value on the perceived distance; through the above adjustment, the model can achieve adaptive control of the individual repulsion behavior with the dynamic change of the emotional state; and then the total formula of the social force model is obtained: ； in, Represents an individual With the wall The forces between them For individuals The speed.

6. The crowd evacuation method based on the coupling of the emotional contagion and social force model according to claim 5, characterized in that, The step 4 comprises the following specific steps: Based on the coupling of social force model and emotion propagation model, the fusion of Path planning and ACO form a multi-agent evacuation simulation framework; The pheromone mechanism in ACO is introduced, and is combined with the local crowd density estimation to simulate the self-organization adjustment of group path selection; the individual continuously releases pheromone to the current grid during the movement, and the system updates the node pheromone value according to the following rules: ； in, For nodes exist The pheromone concentration at any given time; The pheromone evaporation coefficient is used to control its decay rate over time. For the first Individuals at nodes The increase in pheromones left on the surface; Based on the static structure information of the evacuation environment, a shortest path from the current position to the target exit is generated for each individual by the classic algorithm; the path consists of several key navigation nodes, which serve as the initial target sequence of the individual and provide the system with global navigation guidance capability; in the path planning process, to avoid congestion areas, the cost function of the algorithm is improved by introducing a density-aware weight term: ； wherein, is the total cost of the current node , is the cumulative actual cost from the start point to the current node , is the heuristic distance from to the exit, denotes the crowd density at the current node, is the density influence factor for adjusting the weight of local congestion in the path cost. Based on this, a guiding force based on pheromone is added in the social force model: ； where, is the pheromone guidance force, is the gradient of the pheromone field, the direction points to the direction of the fastest increase of concentration; the social force model equation is further changed to: 。 7. The crowd evacuation method based on the coupling of the emotional contagion and social force model according to claim 6, characterized in that, In the step 5, a dynamic path updating mechanism is designed: when the density or pheromone concentration of a certain area exceeds the preset threshold, the individual will trigger local path reconstruction, and the algorithm recalculates the local shortest path from the current position to the exit. The new path also considers the density or pheromone cost, thereby avoiding the impact of local congestion or environmental mutations. In the step 5, a dynamic path updating mechanism is designed: when the density or pheromone concentration of a certain area exceeds the preset threshold, the individual will trigger local path reconstruction, and the algorithm recalculates the local shortest path from the current position to the exit. The new path also considers the density or pheromone cost, thereby avoiding the impact of local congestion or environmental mutations.

Images