A simulation prediction method and system for subway network operation interruption based on digital twin
Through a digital twin method, a digital twin model of urban rail transit system is established to conduct simulation predictions of passenger flow and vehicle flow coupling during subway network operation interruption, which solves the problem of difficulty in integrating complex information in the existing technology and provides comprehensive predictions As a result, it helps dispatchers optimize operational strategies and improve operational efficiency and passenger satisfaction.
Patent Information
- Application Number
- CN202211368300.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-03
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-11-03
AI Technical Summary
It is difficult for the prior art to integrate complex information of the subway system and provide comprehensive and reference-value simulation prediction results for passenger and vehicle flow coupling systems, especially in the event of interruption of subway network operations.
Using a digital twin method, we will establish a digital twin model of the urban rail transit system, obtain basic information of the subway network and train operation data, establish a rule model of response measures, update the network topology model, and conduct effective path search under operation interruption. It also combines the passenger flow management module for passenger flow redistribution and traffic management, and uses the Logit model for passenger flow prediction and traffic control.
Provide comprehensive and reference-worthy simulation solutions and prediction results for the subway dispatch center, assist dispatchers in strategy optimization, improve operational efficiency and passenger satisfaction, and reduce the impact of operational interruptions.
Smart Images

Figure CN115909729B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of smart city rail transit technology and digital modeling technology, and relates to a simulation prediction method and system based on digital twins in the event of a subway network operation interruption. Background Art
[0002] In recent years, rail transit systems have played an increasingly important role in urban transportation systems due to their high capacity, high efficiency, low pollution and low energy consumption. Urban rail construction has continued to grow, and the scale and investment of operation, construction and planning lines have grown by leaps and bounds. Urban rail transit continues to maintain a rapid development trend. Due to the advantages of large capacity, high speed and high punctuality, more and more commuters choose rail transit systems as a means of travel.
[0003] Faced with the growing traffic demand and increasingly serious traffic congestion, the operating pressure of urban rail transit is increasing. In addition, in the networked rail transit, the scope of emergency events is also increasing. Once a risk accident occurs, it may cause train delays at the least, traffic paralysis and congestion at the worst, and may even cause passenger riots, which will have a bad social impact and extremely serious consequences.
[0004] Urban rail transit is a necessity for urban development in the new century, and research on urban rail transit safety management and emergency management systems is particularly important. Rationally allocating and using human and vehicle resources to restore urban rail transit services as soon as possible and reduce adverse effects has become one of the important goals of emergency management.
[0005] The traffic adjustment caused by interruption is relatively complex, and many uncertain factors need to be considered, such as the duration of the interruption, line configuration conditions, and traffic resources, which greatly increases the scope and difficulty of the adjustment. The evaluation of the emergency management plan for traffic interruption requires the simulation and prediction of the flow of people and vehicles during the recovery process of traffic interruption. In this process, the flow of people and vehicles often affects each other. Taking into account the changes in the flow of people and vehicles, using digital twin technology, establishing a rail transit network model, and accurately predicting the process of traffic interruption recovery to support the formulation of emergency management plans has become an important part of rail transit operation management.
[0006] Accurate passenger flow prediction and traffic control are the basis for realizing the operation of urban rail transit systems and optimizing safe dispatching. Only by understanding the passenger flow change rules in advance, grasping the passenger flow characteristics, and accurately predicting the short-term changes in the line network passenger flow after the interruption of driving, can scientific management decisions be made, resources can be reasonably allocated, the impact of failures can be reduced, and passenger satisfaction and operational efficiency can be improved. Summary of the invention
[0007] Technical issues to be solved
[0008] Based on the current research on subway network operation interruption, most of the research focuses on passenger flow prediction or train scheduling plan, and there are few studies that establish models and simulation predictions for passenger flow and vehicle flow coupling systems as a whole. It is difficult to integrate the complex information of the subway system and provide comprehensive and reference-worthy prediction results. In response to the above problems, the present invention provides a simulation prediction method and system based on digital twins in the case of subway network operation interruption.
[0009] Technical Solution
[0010] In order to solve the above problems, the simulation prediction method for subway line network operation interruption based on digital twin provided by the present invention comprises the following steps:
[0011] Step S1: Establish a data interface between the system of the present invention and the digital twin model of the urban rail transit system to obtain basic information of the subway network, train speed, position and other train operation related information; obtain the time, location, expected duration of the interruption event and the direction and degree of its impact on the subway train according to the interruption event information;
[0012] Step S2: For different subway operation interruption scenarios, a rule model of response measures is established, mainly including train operation adjustment rules and passenger flow control rules, and the rules are embedded into the digital twin model of the urban rail transit system;
[0013] Step S3: updating and adjusting the network topology model based on the acquired network data and outage information, and searching for effective paths under operational outages on this basis;
[0014] Step S4: Forecast the passenger flow entering the urban rail transit system under the interruption situation, establish a passenger flow redistribution model under the interruption of subway line network operation, and embed it into the digital twin model of the urban rail transit system to simulate and deduce the passenger flow distribution status.
[0015] Furthermore, in step S1, the basic information of the subway network includes the network topology, the transfer and return functions of the stations, and the section distance information.
[0016] Furthermore, in step S1, the collection of train information includes basic information such as train composition and vehicle capacity, as well as running status information such as real-time position, running speed, and load of the train.
[0017] Furthermore, in step S1, the passenger information collected includes the passenger's identity information, entry station name, entry time, and destination station, and the passenger's exit station name, exit time, and the passenger's intermittent tracking and positioning information are obtained through simulation.
[0018] Furthermore, in step S2, the emergency management plan after the interruption can be input or generated by the system. The method for generating the train route plan is as follows:
[0019] Step S21, update the road network status, find the turnaround station on the line where the interrupted section is located, which is closest to the interrupted section, divide the affected lines into sections (divided into the sum of non-fault sections and faulty sections), and determine the number of interrupted sites and the set of interrupted sites, where the set of interrupted sites constitutes the faulty section, determine the set of passable sites and intervals, and form a new road network structure.
[0020] Step S22: Collect current passenger flow data and calculate the section imbalance coefficient α h , when α h <1.5, the imbalance of passenger flow in the line section is not obvious, and a large route plan can be adopted; when α h When ≥1.5, the passenger flow distribution in the line section is highly unbalanced. In order to better meet the travel needs of passengers, special route plans such as connecting large and small routes or nesting large and small routes should be adopted.
[0021] Furthermore, in step S2, the method for determining the departure time interval is: collecting passenger demand data and vehicle carrying capacity data to calculate the vehicle load factor, using the corrected historical peak hour vehicle load factor as a limiting condition, and through recursive calls, the adjustment and optimization of the departure time interval of the following trains when the line is interrupted can be achieved.
[0022] Furthermore, in step S2, the passenger flow control measures mainly involve station-level passenger flow control. The draft "Urban Rail Transit Operation Safety Management Specifications" stipulates that when the station reaches or exceeds the passenger flow warning line (70% of the carrying capacity), appropriate measures such as flow limitation and station closure should be taken to ensure the safety of passenger transport organization.
[0023] Furthermore, in step S3, the effective path search step is:
[0024] Step S31: determine the constraints for searching for an effective path to ensure that the path is not repeated and to reduce computational redundancy. The constraints are as follows: no repeated sections, no repeated nodes, no failure points, compliance with relative thresholds and absolute thresholds, and a maximum number of transfers of 3 times.
[0025] Step S32: obtaining network interruption data and updating the adjusted network topology model based on the interruption information;
[0026] Step S33: using Dijkstra algorithm to calculate the shortest path between OD pairs and calculate the minimum generalized cost;
[0027] Step S34: Determine all valid paths that meet the search conditions to form a valid path set. Find the associated edges that have not been checked, and dynamically update the function based on the generalized cost and the number of transfers. Each time an edge is included in the valid path set, the generalized cost required for the current path and the number of path transfers are calculated and judged; if the cost exceeds the specified relative threshold, absolute threshold and the number of transfers, the algorithm will no longer traverse the related nodes of the current node, and directly return to the previous layer to search for the next valid path again.
[0028] Furthermore, in step 4, the parameters of the passenger flow redistribution Logit model under the interruption of subway network operation are estimated by piecewise estimation method, and the steps are as follows:
[0029] Step S41: P n (m|t) finds the maximum likelihood function and solves the estimated parameter β m , β tm , m ;
[0030] Step S42: β m , β tm , m Substitute into the formula and solve for V t ;
[0031] Step S43: Set V t P n (t) is the independent variable, and P n (t) Find the maximum likelihood function and solve for the parameter β t , t ;
[0032] Step S44: According to the conditional probability formula, the probability of each selection plan under the influencing factors is solved.
[0033] In the simulation, when passengers choose a specific path, the limitations of passenger flow control measures are also considered, such as the maximum passenger flow entering the station and the passenger flow movement speed.
[0034] Furthermore, the simulation prediction system for subway network operation interruption based on digital twin is characterized by including five modules: basic information management, time management, passenger flow management, vehicle flow management, statistics and visualization:
[0035] The basic information management module manages the basic status parameters of stations and lines, establishes an interface with the database, manages the basic traffic status of the road network, manages the underlying status parameters of the line network after the operation is interrupted, mainly the accessibility of the line and the station, and generates a new route plan after the interruption occurs. The route plan is formulated based on information such as line status and passenger flow. In the simulation, it can be manually input or generated based on this information and the method of formulating the timetable.
[0036] The time management module is mainly responsible for the management of train timetables and the recording of simulation clocks. This module tracks and monitors the running status of subway trains in real time according to the train control system, triggers and records the train's entry, stop, and exit events. When the train is delayed due to various circumstances, small disturbances can be absorbed by adjusting the train speed, while larger disturbances cause operational interruptions, and then switch to the interruption state response. A new planned timetable can be generated after the interruption occurs. This module determines the start and end of the simulation.
[0037] The passenger flow management module includes two parts: passenger flow control measures management and passenger flow forecasting. Passenger flow control measures include closing subway stations, limiting flow, controlling flow, and diverting flow to restrict the flow of passengers entering the station, and record their behavior and location in the subway system, including the time of entering and exiting the station, getting on and off the train, walking in the station, and transferring. The passenger flow forecasting part includes passenger OD pair prediction and passenger flow redistribution. The OD pair prediction generates passenger objects entering the station at a certain time interval based on the influence of passenger flow management measures. The passenger flow forecast is divided into OD pair prediction and passenger flow distribution. After the interruption occurs, all passenger flows can be divided into unaffected passenger flows, lost passenger flows, and diverted passenger flows according to the network interruption situation. Passenger flow forecasts are made under the state of traffic interruption according to different passenger flow types. Comparison of passenger paths under interruption conditions Path plans of various transportation modes (including outside the subway system), and score and probability distribution of each plan.
[0038] The vehicle flow management module collects train operation information from the train control system and manages vehicle information during the simulation process, including location, availability, number of people on board, etc., and updates in real time according to relevant activities. The main vehicle activities include entering and exiting the station, turning back, passing the station, failure, and vehicle detention, among which getting on and off the train, passing the station, and turning back are basic activities, which are applicable in both normal and interrupted situations, while failure and vehicle detention are activities in interrupted scenarios.
[0039] The statistics and visualization module visualizes the complex road network, vehicle flow, and passenger flow information, and displays important indicators in real time through train operation diagrams, heat maps, statistical charts, and other methods. The statistics and visualization indicator module displays the line network status in real time, which is composed of vehicle flow data and passenger flow data. The vehicle flow data is represented by the train operation diagram, the passenger flow is represented by the line network heat map, the number of passengers at the station is represented by the color or size of the point, and the number of passengers in the interval changes with the train, which intuitively displays the real-time simulation situation. The statistical part is centered on four important indicators: the total delay time of the train, the scope of operation adjustment, the delay time of passengers' travel, and the number of passengers affected.
[0040] Beneficial effects of the present invention:
[0041] The present invention is based on the digital twin method, combines the vehicle flow management plan with the passenger flow prediction method, establishes a model for the passenger flow and vehicle flow coupling system as a whole, and simulates and predicts the situation, providing the dispatchers of the subway dispatching center with more comprehensive simulation plans and prediction results with reference value, and assisting the dispatchers in optimizing strategies. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 It is a schematic diagram of the principle of an implementation example of the present invention;
[0043] Figure 2 A flow chart of a method for generating a train timetable and determining a departure time interval according to the present invention;
[0044] Figure 3 Acquisition and service diagram for the digital twin model of the present invention;
[0045] Figure 4 is a flow chart of the path search algorithm of the present invention;
[0046] Figure 5 Schematic diagram of the prediction method of the present invention. DETAILED DESCRIPTION
[0047] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;
[0048] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.
[0049] Step S1: Establish a data interface between the system of the present invention and the digital twin model of the urban rail transit system to obtain basic information of the subway network, train speed, position and other train operation related information; obtain the time, location, expected duration of the interruption event and the direction and degree of its impact on the subway train according to the interruption event information;
[0050] Step S2: For different subway operation interruption scenarios, establish a rule model for response measures, mainly including train operation adjustment rules and passenger flow control rules, and embed the rules into the digital twin model of the urban rail transit system; the acquisition and service process of the digital twin model of the urban rail transit system is based on Figure 3 shown.
[0051] Step S21, update the road network status, find the turnaround station on the line where the interrupted section is located, which is closest to the interrupted section, divide the affected lines into sections (divided into the sum of non-fault sections and faulty sections), and determine the number of interrupted sites and the set of interrupted sites, where the set of interrupted sites constitutes the faulty section, determine the set of passable sites and intervals, and form a new road network structure.
[0052] Step S22: Collect current passenger flow data and calculate the section imbalance coefficient α h , calculated as:
[0053]
[0054] The maximum cross-sectional passenger flow in one direction is A max Indicates that A i is the passenger flow of each station in a one-way section (person / hour), and n is the number of stations along the line. h The closer it is to 1, the more balanced the passenger flow distribution of the maximum cross-section of the line is. h When >1.5, it indicates that the passenger flow imbalance in the line section is serious.
[0055] in accordance with Figure 2 The flowchart of the method for generating the departure time interval of the present invention is shown in FIG. h
[0056] When α h When the ratio is <1.5, the imbalance of passenger flow in the line section is not obvious, and the large route plan can be adopted. This plan is mainly suitable for the situation where the passenger flow of the whole line is relatively even and basically has no difference. The large route plan has simple driving organization, can effectively reduce the number of transfers, and does not need to set up return stations.
[0057] When α h When the ratio is ≥1.5, the passenger flow distribution in the line section is highly unbalanced. In order to better meet the travel needs of passengers, special route plans such as connecting large and small routes or nesting large and small routes should be adopted. By adopting different marshaling and different numbers of vehicles for segmented operation, the transport capacity can be reasonably arranged according to the passenger flow distribution in the section, and the adaptability of the line to the passenger flow demand during the interruption can be improved.
[0058] When the nested route mode of large and small routes is adopted, in order to ensure the train service level and balance the train running time in each section, the departure interval of large route trains is usually an integer multiple of the departure interval of small route trains. If the departure interval of large and small route trains cannot meet this condition, the economic benefits of train operation will be reduced, resulting in an increase in the number of vehicles. Therefore, it is necessary to reasonably formulate a vehicle operation plan under line interruption to meet the maximum passenger travel and evacuation needs with limited vehicle resources.
[0059] In step S2, the method for determining the departure time interval is: collecting passenger demand data and vehicle carrying capacity data to calculate the vehicle full load rate, and using the corrected historical peak hour vehicle full load rate as a limiting condition. The vehicle full load rate refers to the comparison between the maximum cross-sectional passenger flow of the line and the uniform cross-sectional transportation capacity, which is used to measure the vehicle utilization rate. Its formula is expressed as:
[0060]
[0061] Combined with the statistical data of urban rail transit, we can obtain the number of people waiting for urban rail transit during peak hours, the maximum cross-section passenger flow, vehicle full load rate, etc. Figure 2 The train schedule generation flow chart of the present invention shown uses the corrected historical peak hour vehicle load rate as a limiting condition, and through recursive calls, it is possible to adjust and optimize the departure time interval of the following trains when the line is interrupted.
[0062] Step S3: updating and adjusting the network topology model based on the acquired network data and outage information, and searching for effective paths under operational outages on this basis;
[0063] In step S3, the wire mesh modeling method is:
[0064] The urban rail transit network modeling adopts the adjacency matrix representation, which stores the vertex information in the line diagram in a one-dimensional array. The adjacency relationship between the vertices in the diagram is represented by a matrix G = (V, E), where E represents the line in the network, and E = (e1, e2, ...., e M ), M is the number of lines in the network, V i Represents each site in the network, V=(v1,v2,....,v N ), N represents the number of stations, and the section between adjacent stations is represented by S i It means, S=(s1,s2,...,s Q ), Q represents the number of segments in the network, and L represents the line. i It means that L=(l1,l2,...,l M ).
[0065] The adjacent relationship between vertices can be expressed as
[0066]
[0067] When there is an arc between two nodes, the corresponding element is 1, otherwise it is 0.
[0068] Set segment S i ∈SL i , Segment S j ∈SL j , if SL i =SL j Then segment S i and S j Belong to the same line, use 0-1 variable β ij Represents the relationship between the lines belonging to Q sections in the network:
[0069] S i S j ={β ij}
[0070]
[0071] Introducing variable β ij It can effectively determine whether there is a transfer problem on the path. Since a node can belong to multiple routes at the same time, judging whether there is a transfer behavior only by the node is not comprehensive and prone to errors. Therefore, it is necessary to introduce a 0-1 variable β ij By analyzing the relationship between the lines in each section, it can be determined whether transfer behavior occurs.
[0072] The intersection of two lines can be described as line L i With L j , there is a node V i ∈L i And V j ∈L j ,Right now
[0073]
[0074] A transfer node is an intersection of two or more lines in the network. Passengers can complete the transfer between rail lines and reach their destination through the transfer node. Therefore, the number of transfer nodes in the urban rail transit network is an important indicator for measuring the accessibility and complexity of the network. The number of transfer nodes in the network can be represented by an M×M matrix:
[0075] N M×M = {N ij}
[0076]
[0077] Where n ij Indicates line L i With L j The number of transfer nodes.
[0078] In step S3, the effective path search algorithm is:
[0079] Assume that the section is interrupted due to force majeure, and both the up and down sections are not accessible. Determine the stations on both sides of the interrupted section that are suitable for turning back. All trains traveling to the section need to turn back at these two stations using the turning crossovers, and open small routes on both sides of the interrupted area.
[0080] When a train interruption occurs in urban rail transit, the current line network status is updated first. If there is a fault point in the section and no transfer occurs in the section, the section can be directly deleted from the rail network. The section will not be traversed during the path search, thereby satisfying the constraint condition of setting the section as a fault point.
[0081] If the fault point occurs at a node, the node and all sections associated with it are deleted from the rail network to implement the conditional constraint of the fault point.
[0082] The effective path of the urban rail transit network under line interruption is defined as: Let the kth effective path from the starting point O to the end point D be but is the set of nodes of the valid path, and the set of segments that make up the path for represents the generalized cost of k-path, Among them J OD represents the set of all valid paths between ODs, x k represents the number of transfers required for path k. Then path k needs to meet the following conditions:
[0083] The segments are not repeated, that is
[0084] The nodes are not repeated, that is
[0085] There are no failure points. Where V a is the faulty node;
[0086] (4) Relative threshold With absolute threshold Limit the number of valid paths.
[0087] Generally speaking, the relative threshold Take 0.2-1.8, absolute threshold It is more reasonable to take 9-27. This condition limits the effective path to fluctuate within a certain range of the generalized minimum cost path, which can reasonably control the number of effective paths.
[0088] (5) The maximum number of transfers is 3, x k ≤N max ,N max ={2,3}. The number of transfers is the main factor for passengers to make travel choices, so the number of transfers should be introduced as a constraint when searching for effective routes.
[0089] in accordance with Figure 4 The path search algorithm flow chart of the present invention is shown, and the effective path search algorithm flow chart is as follows:
[0090] (1) Import the fault point, delete the corresponding nodes and edges in the graph network, determine the interrupted line, and update the line network data;
[0091] Use Dijkstra algorithm to calculate the shortest path between OD pairs and calculate the minimum generalized cost
[0092] Initialization: v=1, k=1, l=0, c=0;
[0093] Find the associated edge that has not been checked and calculate the generalized cost c of the edge vw , if all the edges associated with the vertex have been checked, the algorithm ends;
[0094] If w is a point that has not been visited, the node type is judged: ① If w is the end point D, the algorithm ends; ② If w is a common node, go to the next step; ③ If w is a transfer node, l = l + 1, and when l ≤ 3, the associated edges are retrieved again;
[0095] The number of transfers and the judgment of generalized costs, if and Store (v,w) into the edge set S, let v=w, k=k+1, and look for unchecked associated edges. If the condition is not met, directly search again.
[0096] When the kth valid path from the starting point O to the end point D is is the set of nodes on the valid path, and the set of segments that make up the path represents the generalized cost of the k-path.
[0097] The Dijkstra algorithm is used to calculate the shortest path between OD pairs. The steps of the algorithm are as follows:
[0098] (1) Let the vertex v1 be labeled P, d(v1) = 0; j T mark d(v j )=l 1j ;
[0099] (2) Take the minimum value among all T labels, such as d(v j0 )=l 1j0 , then v j0 Change the T label of the vertex to P label, and recalculate the T labels of other points with T labels. Select vertex v j T label d(v j ) and d(v j0 )+l j0j Smaller v j New T label for
[0100] (3) Repeat step 2 until v n ∈P, then d(v n ) is the vertex v1 to v n The shortest path length.
[0101] Step S4: Forecast the passenger flow entering the urban rail transit system under the interruption situation, establish a passenger flow redistribution Logit model under the interruption of subway line network operation, and embed it into the digital twin model of the urban rail transit system to simulate and deduce the passenger flow distribution status.
[0102] In step S4, the Logit model can be divided into three parts: determining influencing factors and establishing a utility function; parameter calibration; parameter testing and evaluating the reliability of the model.
[0103] For passenger flows affected by the interruption, the utility function of traveler n choosing option i is
[0104] U in =V in +ε in ;
[0105] The probability that traveler n chooses option i
[0106] P in =Prob(V in +ε in >V jn +ε jn ; i≠j,j∈A n ).
[0107] Among them, V jn is the fixed term in the utility function of traveler n choosing option j, ε jn is the probability term of the utility function of traveler n choosing option j; 0≤P in ≤1, A n Choose a set of options for traveler n.
[0108] Probability term ε in When following independent bi-exponential distributions with uniform parameters, the multinomial logit model is:
[0109]
[0110] Passenger route selection is a two-layer selection tree structure. The upper layer selection set level 2 has T selection options, and the lower layer selection set level 1 has M selection options, where T = (t1, t2, ..., t i ),M=(m1,m2,...,m j ), then the solution selection set A can be expressed as
[0111] A=T×M={(t1×m1),...,(t1×m j ),...,(t2×m1),...,(t2×m j ),(t i×m1),...,(t i ×m j )}
[0112] According to random utility theory, the following assumptions must be met when establishing the traveler utility function: travelers are organisms that make travel choices, and travelers can independently complete travel mode choices without being interfered with by other external factors; travelers will choose the travel plan that best meets their travel needs based on the utility maximization theory.
[0113] The total utility of plan A is
[0114] U in =V t +V m +V tm +ε t +ε m +ε tm ,
[0115] Assuming that the probability terms follow independent bi-exponential distributions with the same parameters, the marginal probability of traveler n choosing option t is
[0116]
[0117] The simulation prediction system for subway line network operation interruption based on digital twin is characterized by including five modules: basic information management, time management, passenger flow management, vehicle flow management, statistics and visualization.
[0118] in accordance with Figure 5 The flowchart of the prediction method of the present invention is shown as follows. After receiving the interruption notification including the interruption time, location, reason and other information, the line network information is updated and the train operation adjustment rules are generated. The time management module generates a new planned train timetable on this basis, and generates a visualized train operation diagram based on computer technology; the vehicle flow management module simulates the activity process of the vehicle flow according to the planned timetable and passenger flow status; the passenger flow management module includes a passenger flow prediction and allocation algorithm, generates inbound passenger flow, and tracks and records passenger flow routes; the statistics and visualization module can display some real-time information in the simulation. After the module simulation is completed, the module records and saves important indicator data from each module and generates a data report.
Claims
1. A simulation prediction method for subway network operation interruption based on digital twin, the method comprising the following steps: Step S1: Establish a data interface between the system and the digital twin model of the urban rail transit system to obtain basic information of the subway network, train running speed, and location train running related information; Obtain the time, location, expected duration of the interruption event and the direction and degree of its impact on subway trains based on the interruption event information; Step S2: For different subway operation interruption scenarios, a rule model of response measures is established, mainly including train operation adjustment rules and passenger flow control rules, and the rules are embedded into the digital twin model of the urban rail transit system; Step S3: updating and adjusting the network topology model based on the acquired network data and outage information, and searching for effective paths under operational outages on this basis; Step S4: Forecast the passenger flow entering the urban rail transit system under the interruption situation, establish a passenger flow redistribution model under the interruption of subway line network operation, and embed it into the digital twin model of the urban rail transit system to simulate and deduce the passenger flow distribution status.
2. The method according to claim 1, characterized in that The rule model of emergency management measures in the scenario of subway operation interruption in step S2 is characterized in that the state-measure-impact rules are embedded in the digital twin model; different management measures and their impacts can be determined by calculation according to different parameter states, or can be directly input.
3. The method according to claim 1, characterized in that The method for updating and adjusting the line network model in step S3 is: obtaining the line network topology model from the digital twin model of the urban rail transit system, and using an adjacency matrix to represent it, storing the vertex information in the line diagram in a one-dimensional array; based on the determined interruption information and importing the fault point, updating and adjusting the line network topology model.
4. The method according to claim 1, characterized in that: The steps of searching for an effective path in step S3 are: Step S31: determining the constraints of the effective path search to ensure that the path is not repeated and reduce computational redundancy, the constraints are as follows: no repeated sections, no repeated nodes, no failure points, meeting the limits of relative thresholds and absolute thresholds, and the maximum number of transfers is 3; Step S32: obtaining network interruption data and updating the adjusted network topology model based on the interruption information; Step S33: Calculate the shortest path between OD pairs and calculate the minimum generalized cost; Step S34: Determine all valid paths that meet the search conditions to form a valid path set.
5. The method according to claim 4, characterized in that In step S4, a basic passenger flow prediction method is used for stations and routes not affected by the interruption event, and an interruption passenger flow prediction method is used for stations and routes affected by the interruption event, and passenger flow is redistributed in combination with the effective path set.
6. A simulation prediction system for subway network operation interruption based on digital twin, characterized in that: It includes five modules: basic information management, time management, passenger flow management, vehicle flow management, statistics and visualization: The basic information management module manages the basic status parameters of stations and lines, establishes an interface with the database, manages the basic traffic status of the line network, and manages the underlying status parameters of the line network after the operation is interrupted, mainly including the accessibility of the line and the station, and the basic train operation rules; The time management module is mainly responsible for the management of train timetables and the recording of simulation clocks. This module tracks and monitors the running status of subway trains in real time according to the train control system, triggers and records the train's entry, stop, and exit events, and can generate a new planned timetable after an interruption occurs. This module determines the start and end of the simulation; The passenger flow management module includes passenger flow control measures and passenger flow prediction. The passenger flow control measures are to close the subway station, limit the flow, control the flow, and divert the flow of passengers entering the station, and record their behavior and location in the subway system, including the time of entering and exiting the station, getting on and off the train, walking in the station, and transferring. The passenger flow prediction part includes passenger OD pair prediction and passenger flow redistribution. The OD pair prediction generates passenger objects entering the station at a certain time interval based on the influence of passenger flow management measures. The passenger flow redistribution is as described in claim 5. The vehicle flow management module collects train operation information from the train control system and manages vehicle information during the simulation process, including location, availability, and number of people on board. It is updated in real time according to relevant activities. The main vehicle activities include entering and leaving the station, turning back, passing the station, failure, and detention; The statistics and visualization module visualizes complex line networks, vehicle flows, and passenger flow information, and displays important indicators in real time through train operation diagrams, heat maps, and statistical charts.
7. The simulation prediction system for subway network operation interruption based on digital twin according to claim 6 is characterized in that After receiving the interruption notification including the interruption time, location and reason information, the line network information is updated and the train operation adjustment rules are generated. The time management module generates a new planned train timetable based on this, and generates a visual train operation diagram based on computer technology; the traffic management module simulates the traffic flow process according to the planned timetable and passenger flow status; the passenger flow management module includes passenger flow prediction and allocation algorithms, generates inbound passenger flow, and tracks and records passenger flow routes; The statistics and visualization module can display some real-time information during the simulation. After the module simulation is completed, the module records and saves important indicator data from each module and generates a data report.
