Dynamic collecting and distributing guiding method for emergency passenger flow evacuation of urban rail transit

Abstract

The invention discloses a dynamic collecting and distributing guiding method for emergency passenger flow evacuation of urban rail transit, which comprises the following steps of: firstly, calculating to obtain a neighborhood radius value meeting service requirements of the number, the position and the like of guiders according to acquired internal construction data of a building, information of the guiders, types of the passengers, real-time position information and relative position information among the passengers; secondly, inputting a neighborhood radius threshold value and a density threshold value, and calculating to obtain the number and positions of clusters after the emergency evacuation passenger flow aggregation and the passenger density in each cluster; calculating to obtain a central aggregation point in each cluster according to the information; and finally, the passenger travels to the aggregation point according to the indication of the mobile phone emergency evacuation guiding software, and meanwhile, the guider arrives at the aggregation point and drives all passengers near the aggregation point to evacuate from the dangerous area in order, so that passenger flow emergency safe evacuation is completed. According to the invention, the operation service level of the urban rail transit system and the travel safety of passengers can be improved when the station is in emergency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of urban rail transit safety and emergency management, and in particular to a dynamic gathering and dispersing guiding method for emergency evacuation of passenger flow in urban rail transit. BACKGROUND

[0002] Urban rail transit stations adopt a three-dimensional spatial layout, the space is relatively closed, the lighting conditions are limited, and it is difficult to evacuate passengers and rescue in the event of various incidents or accidents. The system safety is relatively fragile, and once a fire accident occurs, it may cause casualties and huge property losses, and may even lead to traffic paralysis and affect social stability. Through in-depth research on the behavior of crowd evacuation, evacuation model and simulation, the key nodes in the evacuation process are determined, and the theoretical research including evacuation model construction, emergency environment analysis and evacuation scheme effect evaluation is promoted, which is of great significance to the scientificity of subway emergency plan and the improvement of emergency disposal efficiency.

[0003] How to alleviate and avoid the problems of personnel being trapped and stranded caused by urban rail transit emergencies or accidents, and how to effectively evacuate subway passengers in the event of an accident, are one of the hot topics in current academic research. Through the summary and summary of existing research, it is found that the current research on urban rail transit emergency safety evacuation mainly focuses on evacuation theory research, evacuation method research and evacuation modeling and simulation research.

[0004] There has been sufficient research on the theory and method of urban rail transit emergency passenger flow evacuation, but existing research often focuses on generating evacuation methods through modeling and simulation means for subway operation control system and staff reference, but passengers cannot know the actual evacuation scheme calculated and simulated, and it is difficult to transfer the generated unified or targeted evacuation methods and routes to each passenger in real time and accurately, and the actual implementation effect of the scheme may not meet the expectations.

[0005] When an emergency or accident occurs, the relevant research is generally to explore the direction and path of direct evacuation of passenger flow, but passengers are difficult to calmly face and evacuate steadily according to the scheme instructions in the face of emergency situations, and it is easy to cause congestion and confusion without unified command or guidance.

[0006] Existing research has gradually realized the guiding role of guiding personnel to the evacuated personnel and has carried out related research, but has not carried out specific research on the number of guiding personnel and the specific guiding positions they should reach. In actual emergency evacuation scenarios, the gathering and dispersing positions of each evacuated passenger should be arranged flexibly according to the number of station staff, i.e. guiding personnel, to fully play the role of guiding personnel and arrange guiding personnel to reach the required specific positions, so as to quickly and efficiently guide the evacuation of passengers in the station. SUMMARY

[0007] The present application aims to design a dynamic gathering and guiding method for emergency evacuation passenger flow of urban rail transit, make up for the deficiency of the existing dynamic gathering and guiding method for emergency evacuation passenger flow of urban rail transit in theory and method, formulate the corresponding passenger gathering and guiding scheme according to the actual situation of passenger flow and its distribution characteristics under different space-time attributes in the rail transit station, help the rail transit operating company to reasonably and efficiently respond to emergencies and risk impact, and has important theoretical and practical significance for further improving the service level of urban rail transit system operation and the safety of passenger travel.

[0008] To achieve the above object, the present application provides a dynamic gathering and guiding method for emergency evacuation passenger flow of urban rail transit, and the specific scheme is as follows:

[0009] (1) When an emergency occurs in the urban rail transit station, the internal structure data of the station, the number and position information of the staff in the station, i.e. the guide, and the type information and real-time position information of the passengers are obtained;

[0010] (2) According to the obtained internal structure data of the building, the number and position information of the guide, the type and real-time position information of the passengers and the relative position information between the passengers, the K-nearest neighbor algorithm (KNN) is used to calculate the neighborhood radius ε value that meets the service requirements of the number and position of the guide;

[0011] (3) According to the obtained real-time position information of the passengers and the relative position information between the passengers, the density-based clustering algorithm (DBSCAN) is used to input the neighborhood radius ε and the density threshold MinPts to calculate the number and position of the clusters after the emergency evacuation passenger flow is aggregated and the passenger density in each cluster;

[0012] (4) According to the obtained guide information, the type and real-time position information of the passengers and the relative position information between the passengers, the K-medoids algorithm is used to calculate the center aggregation point in each cluster;

[0013] (5) The passengers go to the aggregation point according to the indication of the mobile phone emergency evacuation guiding software, at the same time, the guide reaches the aggregation point and leads all the passengers near the aggregation point to orderly evacuate the dangerous area, and completes the emergency safety evacuation of passenger flow.

[0014] Further, in step (1), the internal structure data of the station, the number and position information of the staff in the station, i.e. the guide, and the type information and real-time position information of the passengers are obtained.

[0015] Further, the step (2) described is:

[0016] (4.1) KNN algorithm calculates the relative distance D i (q j ,q p ) between passengers q i and q j by an adaptive method according to the real-time position information of passengers and the relative position information between passengers, that is: Wherein, passengers q i ,q j ∈Q, Q is the set of all emergency evacuation passengers, the position coordinates of q i and q j in the n-dimensional feature space are respectively D p is the relative distance value between passengers, p is the distance metric, which is the parameter of the distance formula, and the numerical value is set according to the actual walking characteristics of passengers;

[0017] (4.2) KNN algorithm generates K-distance relative distance ranking graph and deduces discrete points according to the calculated relative distance D p (q i ,q j ) between each passenger, and uses the forward differential formula to calculate the point with large absolute value of second derivative, that is, the inflection point of relative distance: Wherein, α is the discrete point in the differential formula, h is the step in the differential formula, O(h) is an infinitesimal, f ′ (α) is the first derivative, f ″ (α) is the second derivative, the second derivative value at the inflection point is matched with the obtained guide information, and the neighborhood radius ε value meeting the service requirements such as the number and position of guides is quickly found.

[0018] Further, the DBSCAN algorithm in step (3) calculates, judges and generates passenger aggregation clusters meeting the requirements of neighborhood radius ε and density threshold MinPts by S ε (q i )={q j ∈Q|D p (q i ,q j )≤ε} ε (q i ) is the passenger set in the cluster where q i is located, and the density threshold MinPts represents the number of passengers that must be contained in the ε-neighborhood of each aggregation cluster.

[0019] Furthermore, in step (4):

[0020] (6.1) K-medoids algorithm is passed Calculate, judge and generate the central aggregation point within each passenger aggregation cluster, where C is q i The set of central aggregation points in the cluster, c k q i The central aggregation point within the cluster, c k The position coordinates in the n-dimensional feature space are

[0021] (6.2) Analyze and determine whether there are special passengers in the cluster. If not, calculate the generated c k The central gathering point within the cluster is the point where all passengers gather and are evacuated by the guide. If there is a special passenger in the cluster, the calculated central gathering point is automatically converted to the location of the special passenger. The remaining passengers in the cluster are evacuated by the guide after they reach the special passenger's location.

[0022] (6.3) Since some passengers S ε (q i )≠{q j ∈Q|D p (q i ,q j )>ε}, which are not assigned to clusters by the proposed algorithm, and whether there are outliers outside each cluster is analyzed and judged. If not, the passengers in the cluster gather at the central gathering point and are evacuated by the guide; if there are outliers, q j ∈Q, Calculate the nearest central aggregation point in the cluster to the discrete point, guide the outlier passengers to the central aggregation point, and then guide them to evacuate uniformly. β is the central aggregation point in each cluster, c β The position coordinates in the n-dimensional feature space are

[0023] Furthermore, the mobile phone emergency evacuation guidance software described in step (5) will display guidance instructions on the mobile phone interface according to the identity and location of the passengers in the event of an emergency, guiding the passengers to the designated location; at the same time, the passenger center gathering point that needs to be visited will be displayed on the guide's mobile phone interface, and the guide will guide the passengers to evacuate uniformly after arriving at the designated location.

[0024] The specific embodiments provided by the present invention have the following beneficial effects:

[0025] The emergency evacuation guiding software of the mobile phone can display the evacuation scheme and guiding information on the interface of the mobile phone, so that the passengers can obtain the evacuation method and route and the like in a timely manner, and the passengers can be guided in a timely and accurate manner, and the actual implementation effect of the evacuation scheme can be ensured.

[0026] When an emergency occurs, the passengers can be guided by the emergency evacuation guiding software of the mobile phone to go to the area location set in the scheme for temporary collection, and then be led away in an orderly manner by the guide personnel, so that the disordered and inefficient evacuation phenomenon can be avoided, and the efficiency and safety of the evacuation can be improved.

[0027] The emergency evacuation scheme for the passengers in the station can be generated according to the related information of the guide personnel, the role of the guide personnel can be fully played, and the passengers in the station can be guided in a rapid and efficient manner.

[0028] Other features and advantages of the embodiment disclosed in the embodiment will be described in the following specific embodiments, and will be described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 The flow chart of the passenger flow dynamic collection and distribution of the urban rail transit emergency evacuation of the embodiment of the application is shown in the figure;

[0030] Figure 2 The passenger flow schematic diagram of the urban rail transit station platform of the embodiment of the application is shown in the figure;

[0031] Figure 3 The passenger flow dynamic collection and distribution schematic diagram of the urban rail transit station platform of the embodiment of the application is shown in the figure;

[0032] Figure 4 The relative position distance sorting diagram of the passengers in the urban rail transit station platform of the embodiment of the application is shown in the figure;

[0033] Figure 5 The dynamic aggregation cluster schematic diagram of the passenger flow of the urban rail transit emergency evacuation of the embodiment of the application is shown in the figure;

[0034] Figure 6 The central aggregation point schematic diagram in the aggregation cluster of the passenger flow of the urban rail transit emergency evacuation of the embodiment of the application is shown in the figure;

[0035] Figure 7 The central aggregation point change schematic diagram in the aggregation cluster of the passenger flow of the urban rail transit emergency evacuation of the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0036] The technical scheme of the application will be further described below with reference to the accompanying drawings and specific embodiments:

[0037] Embodiment one

[0038] The urban rail transit emergency evacuation passenger flow dynamic gathering and guiding flow process is shown in FIG. 1, and the specific implementation steps are as follows: Figure 1

[0039] In the case subway station, the station dispatching center has a guide, and the station platform has passengers, as shown in FIG. 2, and the specific parameter values are shown in Table 1. At this time, if a fire, flood, earthquake, large passenger flow or other emergency occurs, the passengers need to be quickly evacuated, and the emergency evacuation guiding software in the mobile phone automatically obtains the building internal structure data of the station, the number and position of the staff (passenger emergency evacuation guide) in the station, and the type information and real-time position information of the passengers through mobile signaling. Figure 2

[0040] Table 1 Parameter values

[0041]

[0042]

[0043] When an emergency occurs, the passengers are guided by the mobile phone emergency evacuation software to first go to the area location set in the scheme for temporary gathering, and then are uniformly and orderly led away by the guide, as shown in FIG. 3. Figure 3

[0044] The urban rail transit emergency evacuation passenger flow dynamic gathering and guiding system quickly calculates the neighborhood radius ε value meeting the service requirements of the number and position of the guides according to the obtained building internal structure data, guide information, passenger type and real-time position information, and the relative position information between the passengers, by using the KNN algorithm:

[0045] The KNN algorithm calculates the relative distance between each passenger according to the obtained real-time position information of the passengers and the relative distance D i (q j , q p ) between the passengers q i and q j , by using the formula , where q i , q j ∈Q, Q is the set of all emergency evacuation passengers, and the position coordinates of q i and q j in the n-dimensional feature space are respectively D p is the relative distance value between the passengers, p is the distance measure, which is the parameter of the distance formula, and the numerical value is set according to the actual walking characteristics of the passengers.

[0046] The KNN algorithm generates a K-distance relative distance ranking diagram according to the calculated relative distance value between each passenger by using an adaptive method, as shown in FIG. 4. Figure 4 ​​​As shown, and discrete point derivation is performed, using Among them, α is the discrete point in the differential formula, h is the step size in the differential formula, O(h) is an infinitesimal quantity, f ′ (α) is the first-order derivative, f ″ (α) is the second-order derivative. The second-order derivative value at the inflection point is matched with the obtained guide information to quickly find the neighborhood radius ε value that meets the service requirements such as the number and location of guides. The ε value is 5, or 5 meters.

[0047] According to the real-time location information of passengers and the relative location information between passengers, the DBSCAN algorithm is used to input the neighborhood radius ε and the density threshold MinPts. ε (q i )={q j ∈Q|D p (q i ,q j )≤ε} calculate, judge and generate the number and location of clusters after emergency evacuation passenger flow aggregation and the passenger density within each cluster that meet the requirements of neighborhood radius ε and density threshold MinPts, such as Figure 5 shown.

[0048] According to the obtained guide information, passenger type and real-time location information and the relative location information between passengers, the K-medoids algorithm is used to Calculate, judge and generate the central aggregation point within each passenger cluster, where C is q i The set of central aggregation points in the cluster, c k q i The central aggregation point within the cluster, c k The position coordinates in the n-dimensional feature space are

[0049] As an optional implementation, in order to ensure that the maximum collective walking distance of passengers is less than The position data within the cluster will be calculated multiple times to obtain one or more central aggregation points, and the optimal solution of passenger aggregation clusters and central aggregation points will be obtained, such as Figure 6 shown.

[0050] As an optional implementation, it is analyzed to determine whether there is a special passenger in the cluster. If not, the generated c k It is the central gathering point in the cluster. After the passengers in the cluster gather at the central gathering point, they will be guided to evacuate uniformly by the guide. If there is a special passenger in the cluster, the calculated central gathering point in the cluster will be automatically converted to the position of the special passenger. After the other passengers in the cluster reach the position of the special passenger, they will be guided to evacuate uniformly by the guide. Figure 7 shown.

[0051] As an optional implementation, by formula S ε (q i ≠{q j ∈Q|D p (q i ,q j )>ε}analysis of whether there are outliers outside the cluster, if not, the passengers in the cluster are aggregated after the center aggregation point is aggregated by the guide staff and is guided to evacuate uniformly; if there are outliers, by formula q j ∈Q, the center aggregation point in the cluster closest to the discrete point is calculated, the outlier passenger is guided to the center aggregation point, and the guide staff is guided to evacuate uniformly, wherein c β is the center aggregation point in each cluster, and the position coordinates of c β in the n-dimensional feature space are

[0052] The mobile phone emergency evacuation guidance software displays the guidance instructions on the mobile phone interface according to the identity and position of the passengers in the event of an emergency, and the passengers go to the aggregation point according to the instructions of the mobile phone emergency evacuation guidance software.

[0053] The center aggregation point of the passengers to be guided is displayed on the mobile phone interface of the guide staff, the guide staff reaches the aggregation point and leads all the passengers near the aggregation point to orderly evacuate the dangerous area, and the passenger flow emergency safety evacuation is completed.

Claims

1. A dynamic gathering and dispersing guiding method for emergency evacuation of passenger flow in urban rail transit, characterized in that, It comprises the following steps: (1) When an emergency occurs in an urban rail transit station, the internal structure data of the station, the number and position information of the staff, i.e. the guides, in the station, and the type information and real-time position information of the passengers are obtained; (2) According to the obtained internal structure data of the building, the number and position information of the guides, the type and real-time position information of the passengers, and the relative position information between the passengers, the K-nearest neighbor algorithm (KNN) is used to calculate the neighborhood radius ε value that meets the service requirements of the number and position of the guides; (3) According to the obtained real-time position information of the passengers and the relative position information between the passengers, the density-based clustering algorithm (DBSCAN) is used to input the neighborhood radius ε and the density threshold MinPts to calculate the number and position of the clusters after the aggregation of the emergency evacuation passenger flow and the passenger density in each cluster; (4) According to the obtained guide information, the type and real-time position information of the passengers, and the relative position information between the passengers, the K-medoids algorithm is used to calculate the center aggregation points in each cluster; (5) The passengers go to the aggregation points according to the instructions of the mobile phone emergency evacuation guide software, while the guides arrive at the aggregation points and lead all the passengers near the aggregation points to orderly evacuate the dangerous area, thus completing the emergency safety evacuation of the passenger flow. 2.The dynamic gathering and distributing guidance method for emergency evacuation passenger flow of urban rail transit according to claim 1, characterized in that, In step (1), the internal structure data of the station, the number and position information of the staff, i.e. the guides, in the station, and the type information and real-time position information of the passengers are obtained. 3.The dynamic gathering and distributing guidance method for emergency evacuation of urban rail transit passenger flow according to claim 1, characterized in that, In step (2), the K-nearest neighbor algorithm (KNN) is used to calculate the neighborhood radius ε value that meets the service requirements of the number and position of the guides. (4.1) The KNN algorithm uses an adaptive method to calculate the passenger q based on the obtained real-time passenger location information and the relative position information between passengers. i and q j The relative distance D p (q i ,q j ),Right now: Among them, passenger q i ,q j ∈Q, Q is the set of all emergency evacuation passengers, q i and q j The position coordinates in the n-dimensional feature space are D p is the relative distance between passengers, p is the distance metric, which is a parameter of the distance formula. The value setting depends on the actual walking characteristics of the passengers; (4.2)KNN algorithm by adaptive method, according to the calculated relative distance D between each passenger p (q i ,q j ), generate K-distance relative distance ranking chart, and carry out discrete point derivation, use forward differential formula to solve the point with larger absolute value of second derivative, that is, the inflection point of relative distance: O(h), wherein, alpha is the discrete point in the differential formula, h is the step in the differential formula, O(h) is an infinitesimal, f ′ (alpha) is the first derivative, f ″ (alpha) is the second derivative, match the second derivative value at the inflection point with the obtained guide information, and quickly find the neighborhood radius epsilon value that meets the service requirements such as the number and position of guides.

4. The dynamic gathering and guiding method for emergency evacuation of urban rail transit passenger flow according to claim 1, characterized in that, The DBSCAN algorithm described in step (3) is performed by S ε (q i ) = {q j ∈ Q | D p (q i , q j ) ≤ ε} computes, judges and generates passenger aggregated clusters that satisfy the requirements of neighborhood radius ε and density threshold MinPts, where S ε (q i ) is the passenger set within the cluster where q i is located, and the density threshold MinPts represents the number of passengers that must be contained in the ε-neighborhood of each aggregated cluster.

5. The dynamic gathering and guiding method for emergency evacuation of urban rail transit passenger flow according to claim 1, characterized in that, In step (4), the K-medoids algorithm is used to calculate the center aggregation points in each cluster. (6.1) The K-medoids algorithm works by calculating, determining, and generating a center aggregate point within each passenger aggregate cluster, where C is the set of center aggregate points within the cluster in which q i is located, c k is the center aggregate point within the cluster in which q i is located, c k is the location coordinate of q (6.2) Analysis and judgment of whether there is a special passenger in the cluster, if not, calculate the generated c k That is, the central aggregation point in the cluster, and the passengers in the cluster are guided to evacuate by the guide after aggregation at the central aggregation point; if there is a special passenger in the cluster, the calculated generated central aggregation point in the cluster is automatically converted to the position of the special passenger, and the remaining passengers in the cluster are guided to evacuate by the guide after going to the position of the special passenger. (6.3) Since some passengers S ε (q i )≠{q j ∈Q|D p (q i ,q j )>ε}, which are not assigned to clusters by the proposed algorithm, and whether there are outliers outside each cluster is analyzed and judged. If not, the passengers in the cluster gather at the central gathering point and are evacuated by the guide; if there are outliers, Calculate the nearest central aggregation point in the cluster to the discrete point, guide the outlier passengers to the central aggregation point, and then guide them to evacuate uniformly. β is the central aggregation point in each cluster, c β The position coordinates in the n-dimensional feature space are 6. The dynamic gathering and guiding method for emergency evacuation of urban rail transit passenger flow according to claim 1, characterized in that, In step (5), the mobile phone emergency evacuation guide software displays the guide instructions on the mobile phone interface according to the identity and position of the passengers in the event of an emergency, guiding the passengers to the designated position; meanwhile, the mobile phone interface of the guides displays the center aggregation points of the passengers that need to be visited, and the guides arrive at the designated position to guide the passengers to evacuate.

Images