A simulation method for evacuating people carrying luggage

Abstract

The present invention relates to an evacuation simulation method for a crowd carrying luggage, comprising the following steps: S1: modeling a scene; S2: allocating pedestrians: dividing the scene into grids, and allocating and numbering pedestrians in the grids, including allocating the initial position of the pedestrians and whether the pedestrians carry luggage; S3: calculating the force exerted on each pedestrian in the scene, and determining the expected direction of the pedestrian: the force includes a self-driving force D, a force between pedestrians, and an interaction force between pedestrians and an environmental boundary; S4: calculating a probability matrix, and determining the direction of movement of the pedestrian; S5: updating the position of the pedestrians: updating the position of each pedestrian according to the direction of movement of each pedestrian, the expected speed, and the situation of other people around the pedestrian; S6: stopping the calculation if all pedestrians pass through an exit, otherwise returning to step S3.

Description

Technical Field

[0001] The present invention relates to the technical field of personnel evacuation, and in particular to a personnel evacuation method and system taking into account the influence of luggage. Background Art

[0002] Public transportation hubs such as airports, subways, and train stations have large passenger flows, and pedestrians carrying luggage can be seen everywhere. The luggage carried by passengers in crowded spaces not only takes up a certain amount of available space, but also slows down walking speed and changes movement behavior. Therefore, in order to address the safety management issues in public transportation hubs, it is of great practical significance to develop crowd evacuation technology prediction technology that takes into account the special characteristics of crowds carrying luggage.

[0003] The use of mathematical models to simulate the movement and evacuation process of heterogeneous crowds with luggage has important application value. At present, the research methods for this type of people mainly include: based on the social force model, by increasing the body size and body mass of pedestrians carrying luggage, simulation analysis is carried out in a simple scenario of a straight channel; based on the cellular automaton personnel evacuation model, the space is discretized into grids with different attributes to represent obstacles, pedestrians or exits, and the movement direction of pedestrians is determined by calculating the probability priority matrix of the surrounding neighborhood; the multi-grid model is adopted, that is, the concept of physical force is introduced into the traditional cellular automaton model, by increasing the friction, repulsion and attraction between obstacles and pedestrians, or further adding individual inertia, exit direction preference force, following behavior attraction and other factors, but it is not well applicable to the movement behavior of people carrying luggage. In general, there are still few studies on the model of people carrying luggage, and the existing technology has not yet explored the intrinsic influence mechanism of people carrying luggage on pedestrian movement.

[0004] Due to these problems, the existing models cannot accurately simulate the movement of people carrying luggage. In order to solve the above problems, the present invention fully considers the heterogeneity of pedestrians carrying luggage, establishes an evacuation model that comprehensively considers the impact of pedestrians carrying luggage, and develops a simulation method and system suitable for people carrying luggage, which can realize the simulation of the movement process of people carrying luggage, thereby providing technical support for the fire safety assessment of buildings. Summary of the invention

[0005] The present invention provides a method for simulating the evacuation of a crowd carrying luggage, comprising the following steps:

[0006] S1: Modeling the scene: spatial modeling, marking features related to evacuation, including the spatial structure of the building, walls, obstacles, and exit locations;

[0007] S2: Assign pedestrians: Divide the scene into grids, assign pedestrians to the grids and number them, including assigning the initial positions of pedestrians and whether the pedestrians are carrying luggage;

[0008] S3: Calculate the force on each pedestrian in the scene and determine the pedestrian's desired direction: the force includes the self-driving force D, the force between pedestrians, and the interaction force between pedestrians and the environment boundary;

[0009] S4: Calculate the probability matrix and determine the pedestrian's movement direction: Calculate the probability matrix according to the force applied to each pedestrian, and determine the movement direction of each pedestrian according to the maximum probability in the probability matrix;

[0010] S5: Update pedestrian position: Update the position of each pedestrian according to the direction of movement, expected speed, and other people around the pedestrian;

[0011] S6: If all pedestrians pass through the exit, stop the calculation, otherwise return to step S3.

[0012] Furthermore, in step S3, the magnitude of the self-driving force of each pedestrian is set between 0 and 1, and the direction of the self-driving force D is determined by the maximum probability value of the pedestrian moving towards the exit n.

[0013] Furthermore, the interaction force between individuals on pedestrian i includes the interaction force of pedestrian j on pedestrian i and the non-contact force of pedestrian j on pedestrian i.

[0014] Furthermore, the interaction force between the pedestrian i and the environment boundary includes a contact force and a non-contact force between the pedestrian i and the environment boundary.

[0015] Furthermore, the pedestrian's movement direction is calculated by the probability matrix P a,b We get: (a, b) represents the direction of the eight grids around pedestrian i, and the movement direction of pedestrian i at the next moment is the direction of the maximum value of the probability matrix in the eight grid directions around pedestrian i.

[0016] Furthermore, the next step position of pedestrian i is determined by the following formula:

[0017] V i =min{a i d ij -r ij ,L d,i},

[0018] Where V i represents the number of grids where pedestrian i is moving, d ij represents the number of grids between pedestrian i and the nearest pedestrian j in the direction of motion, a i is the corresponding weight coefficient, rij is the sum of the body radii of pedestrian i and pedestrian j, L d,i It represents the maximum number of grids that pedestrian i can move when reaching the desired speed.

[0019] The present invention quantitatively summarizes the interaction between pedestrians carrying luggage and moving people from the physical and psychological levels, and can restore the interaction mechanism between pedestrians carrying luggage and other people. At the same time, the classic multi-grid model is modified from the influence of body shape, interaction force between individuals and individual movement speed, and the movement characteristics of people carrying luggage can be simulated. It has great research significance and practical application value for the reasonable setting of spatial layout and evacuation channels of public places, scientific formulation of emergency response plans, evacuation strategies and rescue measures, ensuring the safety of people's lives and reducing the impact of accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a main flow chart of the evacuation simulation method for people carrying luggage of the present invention;

[0021] Figure 2 is a schematic diagram of the direction selection angles of a pedestrian carrying a suitcase;

[0022] Figure 3 It is a schematic diagram of the decomposition of the self-driving force;

[0023] Figure 4 is a schematic diagram of the distance between pedestrians i and j;

[0024] Figure 5 It is a schematic diagram of the decomposition of the interaction forces between pedestrians;

[0025] Figure 6 is a schematic diagram of the distance between pedestrian i and the boundary;

[0026] Figure 7 It is a schematic diagram of the probability priority matrix;

[0027] Figure 8 It is a schematic diagram of the calculation scenario;

[0028] Fig. 9 It is the trajectory diagram of people carrying suitcases in the straight channel. DETAILED DESCRIPTION

[0029] See also Figure 1 The evacuation simulation method for people carrying luggage of the present invention mainly comprises the following steps:

[0030] S1: Model the scene and mark the features related to evacuation, including the spatial structure of the building, walls, obstacles, and exit locations;

[0031] S2: Assign pedestrians: Divide the scene into grids, assign pedestrians to the grids and number them, including assigning the initial positions of pedestrians and whether the pedestrians are carrying luggage;

[0032] S3: Calculate the forces on each pedestrian in the scene, including the self-driving force D, the force between pedestrians, and the interaction force between pedestrians and the environment boundary;

[0033] S4: Calculate the probability matrix based on the force applied to each pedestrian and determine the movement direction of each pedestrian;

[0034] S5: Determine the next position of each pedestrian based on the direction of movement, expected speed, and other people around the pedestrian;

[0035] S6: If all pedestrians pass through the exit, stop the calculation, otherwise return to step S3.

[0036] In the simulation process, the focus is on determining the next movement direction of each pedestrian. In the present invention, the control of the movement direction mainly considers the influence of surrounding pedestrians and obstacles, and is mainly determined by the probability priority matrix of three forces: self-driving force, interaction force between individuals, and interaction force between individuals and environmental boundaries.

[0037] The scene is divided into grids, each pedestrian is in a grid, and each pedestrian is surrounded by eight grids. All pedestrians are numbered, and for each pedestrian i, the force it is subjected to is introduced as follows:

[0038] 1. Self-driving force D

[0039] The self-driving force indicates the purposefulness of pedestrian movement or the expected degree of movement. Its calculation method refers to the classic multi-grid model (XU X, SONG WG, ZHENG H Y. Discretization effect in a multi-grid egress model [J]. Physica A: Statistical Mechanics and its Applications, 2008, 387 (22): 5567-74.): The size of the self-driving force can be set according to the pedestrian's expected degree of movement to the destination, and the value range is between 0 and 1.

[0040] The direction of the self-driving force is the pedestrian's desired direction. The pedestrian's desired direction selection is a complex behavior coupled with multiple factors. In addition to considering the influence of factors such as the distance from the exit, the crowd density near the exit, and the width of the exit, pedestrians carrying suitcases will also make dynamic decisions on the desired direction based on the situation of carrying their own suitcases. In the model, combined with the characteristics of pedestrians carrying suitcases, the desired direction calculation module mainly considers four aspects: the distance from the exit, the crowd density near the exit, the width of the exit, and the pedestrian's direction selection angle. The details are as follows:

[0041] (1) The distance from pedestrians to the exit. The closer the exit is to the pedestrians, the more likely it is to be selected.

[0042] The greater the rate, the greater the distance d from pedestrian i to exit n. i,n for:

[0043]

[0044] Among them, (x i ,y i ) is the coordinates of the center grid point of pedestrian i, (x n ,y n ) is the coordinate of the midpoint of exit n.

[0045] (2) Crowd density near the exit. When the density of pedestrians near one exit is high, other pedestrians are likely to give up choosing this exit. Define the local density ρ near exit n n for:

[0046]

[0047] Where N is the number of pedestrians near the exit, w n represents the actual width of exit n, λ is a parameter used to adjust the influence range of exit n, and ω is a parameter used to adjust the influence of pedestrians carrying luggage on the crowd density.

[0048] (3) Exit characteristic width. In the calculation of the expected direction of ordinary pedestrians, the larger the exit width, the greater the probability that the exit will be selected. Among people carrying suitcases, considering that the presence of suitcases will affect pedestrians' judgment of the width allowed to pass through the exit, for pedestrians carrying suitcases, the improved ratio of exit width to pedestrian characteristic parameters is used to reflect the impact of exit width on the expected direction selection of people carrying suitcases:

[0049]

[0050] where w′ i,nis the characteristic width of exit n for different types of pedestrians, μ is a parameter used to reflect the characteristics of individual types. For ordinary pedestrians, μ=1. For pedestrians carrying suitcases, μ takes a value between 1.1 and 2, which is related to individual characteristics and the size of the suitcase.

[0051] (4) Direction selection angle. When a pedestrian needs to make a direction choice, he or she is more likely to make a small change of direction. Define the direction selection angle θ of pedestrian i to exit n as i,n is the angle between the line connecting pedestrian i and exit n at the current moment and the velocity direction vector. Figure 2 As shown, it is a schematic diagram of the direction selection angle of a pedestrian carrying a suitcase.

[0052] On this basis, the probability P of pedestrians choosing exit n in the expected direction calculation module is obtained: i,n , the pedestrian’s expected direction in the next time step points to the exit with the highest probability of selection.

[0053]

[0054] The probability calculation formula for each influencing factor is:

[0055]

[0056]

[0057] The corresponding weight coefficient calculation formula is:

[0058]

[0059] Where N is the number of exports, P i,n (d), P i,n (ρ), P i,n (w), P i,n (θ) represent the probability of pedestrians choosing each exit when only considering the distance from the pedestrian to the exit, the density near the exit, the exit width and the direction selection angle. i,n (ρ) is actually irrelevant to the specific pedestrians but is related to the density of people near the exit. To maintain consistency, the subscript i is also added. d , k ρ , k w , k θ They are the corresponding adjustment factors used to adjust the sensitivity of pedestrians to the four influencing factors. α, β, γ, are the weight coefficients of distance, local density, exit width and direction selection angle, k α , k β , k γ , There are four adjustment factors, which are used to adjust the sensitivity of the four weights.

[0060] Since the model adopts a discretized eight-direction walking rule, the direction of the self-driving force does not always coincide with one of the eight directions. Therefore, the self-driving force is projected to the three directions closest to it according to the force decomposition method, such as Figure 3 As shown, the self-driving force D is projected to the three directions of (-x, 0), (-x, y) and (0, y), respectively. -x,0 ,D -x,y ,D 0,y Represents the projection value of D in these three directions.

[0061] 2. Interaction forces between individuals

[0062] Referring to the classic multi-grid model, the influence of luggage on the spatial distribution of pedestrians during movement and the interaction between individuals are considered. Contact force between pedestrians and non-contact forces It consists of two parts (where the subscript ij represents the interaction between pedestrians i and j), and the specific formula is as follows:

[0063]

[0064] The contact force represents the squeezing, repulsion and friction between pedestrians, including the contact force between ordinary pedestrians, between ordinary pedestrians and pedestrians carrying luggage, and when pedestrians carrying luggage overlap. The contact force calculation method in the present invention is as follows:

[0065]

[0066] where F ij represents the average mutual squeezing effect between pedestrians, θ is F ij The angle between the eight directions of movement of pedestrians in the model is between -180° and 180°, with counterclockwise direction as positive and clockwise direction as negative. ij Represents the friction coefficient, which has different values ​​for different types of contact (see Figure 5 ).

[0067] Based on experimental research results, pedestrians carrying suitcases tend to keep a greater distance from the surrounding people and environmental boundaries. In order to reflect this feature, for pedestrian i carrying a suitcase, a non-contact force is added to the interaction force between individuals. The specific formula is as follows:

[0068]

[0069] Where A and B are the intensity parameter and characteristic distance parameter of the non-contact force, respectively.ij represents the sum of the body radii of pedestrian i carrying a suitcase and other pedestrians j, d ij represents the distance between the center grid points of pedestrian i and pedestrian j carrying a suitcase, such as Figure 4 shown.

[0070] 3. The interaction between individual and environmental boundaries

[0071] Interaction between individual and environmental boundaries The calculation method of is similar to the calculation method of the interaction between individuals, which is also divided into the contact force between pedestrians and the environmental boundary and the non-contact forces between the pedestrian carrying the suitcase and the environmental boundary The two parts are calculated as follows:

[0072]

[0073] where F iW It is used to represent the average mutual squeezing effect between pedestrians or luggage and the environment boundary, θ is F iw The angle between the eight directions of movement of pedestrians in the model is between -180° and 180°, with counterclockwise direction as positive and clockwise direction as negative. iw represents the friction coefficient, A and B are the intensity parameter and characteristic distance parameter of the non-contact force, respectively, and r i represents the body radius of pedestrian i, d iw Represents the distance between pedestrian i and the environment boundary, and its schematic diagram is as follows Figure 6 shown.

[0074] In step S4, the movement direction of the pedestrian individual is calculated by calculating the probability matrix P a,b get:

[0075] By summing the forces exerted on pedestrian i by all pedestrians (represented by peds in the formula) within a certain radius (generally 2m) around pedestrian i, the description of the pedestrian's movement direction is obtained, as follows:

[0076]

[0077] See also Figure 7 , each square around a pedestrian represents 8 directions, which are represented by (a, b) here. a,b It is used to reflect whether the movement of the pedestrian to the position (a, b) violates the volume exclusion principle, that is, whether the pedestrian will overlap with other pedestrians or walls if he moves to the position (a, b). If it does not violate the principle (no overlap occurs), it is 1, otherwise it is 0.

[0078] I a,bIt is the inertia factor, indicating that the pedestrian has a tendency to continue moving in the direction of the previous step. If the direction of the position (a, b) relative to the pedestrian is the same as the direction of the pedestrian's movement in the previous time step, it is greater than 1, otherwise it is 1. a,b is the projection value of the self-driving force D in the (a, b) direction, and They represent the interaction between pedestrians and the interaction between pedestrians and the environment boundary, respectively, and the normalization factor Used to ensure that P calculated in 8 directions a,b The sum is 1. By calculating the 8 probability values ​​in the Moore neighborhood around the pedestrian, such as Figure 7 As shown, the probability priority matrix of the pedestrian at the current moment can be obtained, and the grid with the largest probability is selected to determine the movement direction of the pedestrian at the next moment.

[0079] After the direction of movement of pedestrian i is determined, the next step is to determine the speed of the pedestrian. The speed of movement mainly considers the influence of expected speed and crowd density, and is mainly determined by two factors: individual expected speed and the nearest neighbor distance in the direction of movement.

[0080] There is a close relationship between individual movement speed and forward distance. As the forward distance between individuals increases, the movement speed of pedestrians will also increase. In the model, we consider the influence of pedestrians' expected speed and forward distance, and improve the calculation method of individual movement speed in the multi-grid model, as shown in formula (6).

[0081] V i =min{a i d ij -r ij ,L d,i} (6)

[0082] Among them, d ij represents the number of grids between pedestrian i and the nearest pedestrian j in the direction of motion, a i is the corresponding weight coefficient, r ij is the sum of the body radii of pedestrian i and pedestrian j, L d,i It represents the maximum number of grids that pedestrian i can move when reaching the desired speed. It can be set differently depending on whether the pedestrian is carrying a suitcase.

[0083] The following is a description of the calculation method of the model with a specific example:

[0084] Step 1: Build the scene through coordinate data and initialize crowd information. According to the simulation scene, determine the coordinate system and build the building boundary ( Figure 8), the width of the straight channel is 3m and the length is 10m. The length direction is the x direction and the width is the y direction. The initial position of the crowd is 4.2m away from the channel. There are 68 people in total, arranged in rows of 4 people. The proportion of people carrying suitcases is 50%, and they are evenly distributed in the crowd. The body size of the individual is r i The suitcase is rectangular, with a length and width of l lug and w lug , the specific parameters are shown in Table 1.

[0085] Table 1 Summary of evacuation motion model parameters

[0086]

[0087] Step 2: After determining the spatial distribution information of the scene and the initial position information of the personnel, the probability of each person choosing each exit can be calculated according to formula (4): i,n .

[0088] Step 3: Calculate the probability priority matrix of each person's grid to the adjacent grid according to formula (5), so as to determine the direction of the person at the next moment.

[0089] Step 4: Calculate the next position of the person according to formula (6).

[0090] Step 5: If all people have passed through the exit, stop the calculation, otherwise continue to simulate the pedestrians in the scene.

[0091] The final simulated evacuation trajectory is as follows: Fig. 9 shown.

[0092] The present invention quantitatively summarizes the interaction between pedestrians carrying luggage and moving people from the physical and psychological levels, and can restore the interaction mechanism between pedestrians carrying luggage and other people. At the same time, the classic multi-grid model is modified from the influence of body shape, interaction force between individuals and individual movement speed, and the movement characteristics of people carrying luggage can be simulated. It has great research significance and practical application value for the reasonable setting of spatial layout and evacuation channels of public places, scientific formulation of emergency response plans, evacuation strategies and rescue measures, ensuring the safety of people's lives and reducing the impact of accidents.

[0093] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for simulating the evacuation of people carrying luggage, characterized in that: The following steps are involved: S1: Modeling the scene: spatial modeling, marking features related to evacuation, including the spatial structure of the building, walls, obstacles, and exit locations; S2: Assign pedestrians: Divide the scene into grids, assign pedestrians to the grids and number them, including assigning the initial positions of pedestrians and whether the pedestrians are carrying luggage; S3: Calculate the force on each pedestrian in the scene and determine the pedestrian's desired direction: the force on the pedestrian includes the self-driving force D, the force between pedestrians, and the interaction force between pedestrians and the environment boundary; S4: Calculate the probability matrix and determine the pedestrian's movement direction: Calculate the probability matrix according to the force applied to each pedestrian, and determine the movement direction of each pedestrian according to the maximum probability in the probability matrix; S5: Update pedestrian position: Update the position of each pedestrian according to the direction of movement, expected speed, and other people around the pedestrian; S6: If all pedestrians pass through the exit, stop the calculation, otherwise return to step S3; In step S3, the magnitude of each pedestrian's self-driving force is set between 0 and 1, and the direction of the self-driving force D is determined by the maximum probability value of the pedestrian moving toward exit n. i The probability of moving to exit n Determined by the following formula: ， in, , , , Pedestrians i Distance to the exit, density of people at the exit, width of the exit, pedestrians i and exit directions for pedestrians i The probability of moving to exit n, is the corresponding weight coefficient; , , , , , is the corresponding adjustment factor, For pedestrians i The distance to exit n, N is the number of pedestrians near the exit, is the density of people at the exit, is the characteristic width of outlet n, For pedestrians i The direction to exit n is selected at an angle; specifically: , represents the actual width of outlet n, is the parameter that characterizes the impact of exports. is a parameter used to adjust the effect of pedestrians carrying luggage on crowd density. , is a parameter used to reflect the characteristics of individual types. For ordinary pedestrians, , for pedestrians carrying suitcases, The value is between 1.1 and 2; In step S4, the pedestrian's movement direction is calculated by calculating the probability matrix get: ， Where (a, b) represents pedestrians i The directions of the eight surrounding grids, is the normalization factor, ensuring that the P calculated in 8 directions is a,b The sum is 1; It is used to reflect whether the pedestrian's movement to position (a, b) violates the volume exclusion principle, and the value is 0 or 1; is the inertia factor, which indicates that the pedestrian i Has the tendency to continue moving in the direction of the previous step; is the projection value of the self-driving force D in the (a, b) direction, and peds represents pedestrians i The number of other pedestrians within a certain range; i The direction of movement at the next moment is the pedestrian i The maximum direction of the probability matrix in the eight surrounding grid directions; In step S5, the pedestrian i The next step position is determined by: ， in represents the number of grids where pedestrian i is moving, Indicates pedestrian i The closest pedestrian in the direction of movement j The number of grids between is the corresponding weight coefficient, It's a pedestrian i With pedestrians j The sum of the body radii, Indicates pedestrian i The maximum number of grid cells that can be moved while achieving the desired speed.

2. The method according to claim 1, characterized in that: In step S3, the pedestrian i The interaction force between the individuals is expressed as: ， in Indicates pedestrian j For pedestrians i The interaction force, For pedestrians j For pedestrians i The contact force, For pedestrians j For pedestrians i The non-contact force indicates that the pedestrian i When carrying a suitcase, be careful with pedestrians. j The forces between them; ， Where A and B are the intensity parameter and characteristic distance parameter of the non-contact force, respectively. Pedestrians carrying suitcases i and other pedestrians j The sum of the body radii, Pedestrians carrying suitcases i and pedestrians j The distance between the center grid points.

3. The method according to claim 2, characterized in that In step S3, the pedestrian i The interaction force with the environment boundary is expressed as: ， in Indicates pedestrian i The forces acting on the environment boundaries, For pedestrians i The contact force with the environment boundary, Indicates pedestrian i Non-contact forces with the environmental boundaries when carrying a suitcase ; ， Indicates pedestrian i The body radius, Indicates pedestrian i The distance between the environment boundary W.

Images