Urban rail transit operation scheduling method and system

Abstract

The present invention relates to the technical field of rail transit operation and scheduling, and discloses an urban rail transit operation and scheduling method and system, wherein the urban rail transit operation and scheduling method comprises the following steps: collecting rail environment information; generating No. 1 two-dimensional structural data from the collected rail environment information, wherein the No. 1 two-dimensional structural data comprises No. 1 data matrix and No. 1 relationship matrix; inputting the No. 1 two-dimensional structural data, a population distribution map of the rail operation area, and a ground road flow map of the rail operation area into a rail transit operation and scheduling model, and the rail transit operation and scheduling model outputs a scheduling strategy; the present invention can comprehensively consider rail transit scheduling during an earthquake by integrating non-rail system information such as ground transportation, personnel distribution, and shock-shielding zone distribution, so as to enable rail vehicle passengers to obtain better shelter and minimize the economic losses caused by the earthquake.

Description

Technical Field

[0001] The present invention relates to the technical field of rail transit operation scheduling, and more specifically, to an urban rail transit operation scheduling method and system. Background Art

[0002] Urban rail transit systems, as crucial transportation infrastructure in large cities, face severe operational challenges during natural disasters such as earthquakes. A rational and efficient dispatching plan is crucial to maximizing passenger and staff safety and minimizing property losses. However, complex dynamic events like earthquakes, which cause rapid changes in operating conditions such as network topology and transport resources, pose significant challenges to traditional dispatching optimization methods. Summary of the Invention

[0003] The present invention provides an urban rail transit operation scheduling method and system to solve technical problems in related technologies.

[0004] The present invention provides an urban rail transit operation scheduling method, comprising the following steps:

[0005] Step 100: Collecting rail environment information, including earthquake data, population distribution map of the rail operation area, ground road traffic map of the rail operation area, earthquake safety zone map of the rail operation area, rail transportation network topology map, and capacity and status data of all rail vehicles;

[0006] Step 200: Generate first two-dimensional structural data from the collected track environment information. The first two-dimensional structural data includes a first data matrix and a first relationship matrix. Each cell of the first data matrix represents the first one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transit network topology map, shock-proof safety areas, and earthquake sources. This means that a cell of the first data matrix only contains the track environment information of the independent object it represents.

[0007] The element in the i-th row and j-th column of the No. 1 relationship matrix represents the association between the i-th unit and the independent object represented by the j-th unit of the No. 1 data matrix. If there is an association, the value of this element in the No. 1 relationship matrix is ​​1, otherwise it is 0;

[0008] A number one one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the orbital environment information of the corresponding independent object collected at the t-th moment;

[0009] Step 300: Input the first two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolutional layer, a second convolutional layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolutional layer and the second convolutional layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer.

[0010] The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers.

[0011] Step 400: Send the dispatching strategy to the rail transit dispatching center for execution to control the operation of rail vehicles.

[0012] Furthermore, the existence of a correlation between rail vehicles means that: the rail vehicles are on the same track;

[0013] There is a relationship between stations if: the stations are located on the same track;

[0014] The station and the rail vehicle are associated when: the station closest to the rail vehicle is the station;

[0015] The relationship between the earthquake source rail vehicle and the shock-absorbing safety area means that the rail vehicle and the shock-absorbing safety area are located within the destructive influence range of the earthquake.

[0016] Furthermore, the orbital environment information of the earthquake source is earthquake data.

[0017] Furthermore, the calculation formula of the first intermediate layer is as follows:

[0018] u (t) =σ(W u X (t) +W u H (t-1) +b u )

[0019] r (t) =σ(W r X (t) +W r H (t-1) +b r )

[0020] C (t) =tanh(W c X(t) +W c r (t) ⊙H (t-1) +b c )

[0021] H (t) =(1-u (t) )⊙C (t) +u (t) ⊙H (t-1)

[0022] Among them, W u 、W r 、W c represents the weight parameter, b u 、b r 、b c represents the bias parameter, ⊙ represents the dot product, u (t) 、r (t) and C (t) Represent the first, second, and third intermediate states respectively, where X (t) Represents the tth data item of one-dimensional structure data, H (t) and H (t-1) They represent the t-th and t-1-th intermediate representation data respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1, H (t-1) =X (t) , tanh is the hyperbolic tangent function, and σ represents the sigmoid function.

[0023] Furthermore, the calculation formula of the second intermediate layer is as follows:

[0024]

[0025] Where kv represents the second intermediate representation data of the v-th unit of the data matrix No. 1, M(v) is the set of units associated with the v-th unit of the data matrix No. 1, The nth first intermediate representation data output when the first one-dimensional structure data of the uth unit of the first data matrix is ​​input into the first intermediate layer, c v Is a normalization constant, the default value is M (v) The total number of units in , σ is the sigmoid function, W k Represents the third weight parameter.

[0026] Furthermore, the calculation formula of the first output layer is as follows:

[0027]

[0028] Where y represents the first output vector, its cth component value represents the probability value of the cth scheduling strategy, and the scheduling strategy with the largest probability value is selected as the output. The scheduling strategy group contains all executable scheduling strategies, M(all) represents the set of all units of the data matrix, P1 represents the first image feature, P2 represents the second image feature, and W y represents the output weight parameter, and by represents the output bias parameter.

[0029] Furthermore, the training data of the rail transit operation scheduling model comes from a simulation system. The simulation system simulates the rail transit information that needs to be scheduled and simulates the earthquake source to perform simulation, and obtains training data in a simulation environment.

[0030] Furthermore, the loss function for training the rail transit operation scheduling model is as follows:

[0031]

[0032] SU people and SU property Represents the personnel loss and property loss after the scheduling strategy is executed, people max Indicates the total number of passengers on the rail vehicle, property max Indicates the total assets of the rail transit system, WU people and WU property They represent the reduction in personnel losses and property losses compared to not implementing the scheduling strategy.

[0033] The present invention provides an urban rail transit operation and dispatching system, comprising: a first module for collecting rail environment information, the rail environment information including earthquake data, a population distribution map of the rail operation area, a ground road flow map of the rail operation area, a shock-proof safety area map of the rail operation area, a rail transit network topology map, and the capacity and status data of all rail vehicles;

[0034] The second module generates No. 1 two-dimensional structural data from the collected track environment information. The No. 1 two-dimensional structural data includes No. 1 data matrix and No. 1 relationship matrix. One unit of the No. 1 data matrix represents No. 1 one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transportation network topology map, shock-proof safety areas, and earthquake sources. It means that a unit of the No. 1 data matrix only contains the track environment information of the independent object it represents. The track environment information of the earthquake source is earthquake data.

[0035] The third module inputs the No. 1 two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolution layer, a second convolution layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolution layer and the second convolution layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer;

[0036] The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers.

[0037] The fourth module sends the scheduling strategy to the rail transit scheduling center for execution to control the operation of rail vehicles.

[0038] The present invention provides a computer storage medium for storing computer-readable instructions, which can run the aforementioned one-stop threat analysis and supervision command system when the computer-readable instructions are read.

[0039] The beneficial effects of the present invention are:

[0040] The present invention can comprehensively consider rail transit scheduling during earthquakes by integrating non-railway system information such as ground transportation, personnel distribution, and earthquake-proof zone distribution, so as to enable rail vehicle passengers to get better shelter and minimize the economic losses caused by the earthquake.

[0041] The deep reinforcement learning technology of the present invention generates scheduling decisions directly from high-dimensional environmental states without the need for detailed modeling and can flexibly respond to complex dynamic scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of an urban rail transit operation scheduling method of the present invention;

[0043] Figure 2 It is a module diagram of the rail transit operation scheduling model of the present invention. DETAILED DESCRIPTION

[0044] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that these embodiments are discussed solely to enable those skilled in the art to better understand and implement the subject matter described herein, and that the functions and arrangements of the elements discussed may be varied without departing from the scope of this specification. Various examples may omit, substitute, or add various processes or components as needed. Furthermore, features described in some examples may be combined in other examples.

[0045] At least one embodiment of the present invention discloses a method for scheduling urban rail transit operations. Figure 1 As shown, the following steps are included:

[0046] Step 100: Collecting rail environment information, including earthquake data, population distribution map of the rail operation area, ground road traffic map of the rail operation area, earthquake safety zone map of the rail operation area, rail transportation network topology map, and capacity and status data of all rail vehicles;

[0047] Step 200: Generate first two-dimensional structural data from the collected track environment information. The first two-dimensional structural data includes a first data matrix and a first relationship matrix. Each cell of the first data matrix represents the first one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transit network topology map, shock-absorbing safety areas, and earthquake sources. This indicates that a cell of the first data matrix only contains the track environment information of the independent object it represents. The track environment information of the earthquake source is earthquake data.

[0048] The element in the i-th row and j-th column of the No. 1 relationship matrix represents the association between the i-th unit and the independent object represented by the j-th unit of the No. 1 data matrix. If there is an association, the value of this element in the No. 1 relationship matrix is ​​1, otherwise it is 0;

[0049] There is a relationship between rail vehicles if: the rail vehicles are on the same track;

[0050] There is a relationship between stations if: the stations are located on the same track;

[0051] The station and the rail vehicle are associated when: the station closest to the rail vehicle is the station;

[0052] The relationship between the earthquake source rail vehicle and the earthquake-proof safety area means that the rail vehicle and the earthquake-proof safety area are located within the destructive influence range of the earthquake;

[0053] A number one one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the orbital environment information of the corresponding independent object collected at the t-th moment;

[0054] Step 300: Input the first two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolutional layer, a second convolutional layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolutional layer and the second convolutional layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer.

[0055] The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers.

[0056] Step 400: Send the dispatching strategy to the rail transit dispatching center for execution to control the operation of rail vehicles.

[0057] At least one embodiment of the present invention discloses an urban rail transit operation dispatching system, comprising:

[0058] The first module collects rail environment information, including earthquake data, population distribution maps of the rail operation area, ground road traffic maps of the rail operation area, earthquake safety zone maps of the rail operation area, rail transportation network topology maps, and capacity and status data of all rail vehicles;

[0059] The second module generates No. 1 two-dimensional structural data from the collected track environment information. The No. 1 two-dimensional structural data includes No. 1 data matrix and No. 1 relationship matrix. One unit of the No. 1 data matrix represents No. 1 one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transportation network topology map, shock-proof safety areas, and earthquake sources. It means that a unit of the No. 1 data matrix only contains the track environment information of the independent object it represents. The track environment information of the earthquake source is earthquake data.

[0060] The third module inputs the No. 1 two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolution layer, a second convolution layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolution layer and the second convolution layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer;

[0061] The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers.

[0062] The fourth module sends the scheduling strategy to the rail transit scheduling center for execution to control the operation of rail vehicles.

[0063] At least one embodiment of the present invention discloses a computer storage medium for storing computer-readable instructions. When the computer-readable instructions are read, the aforementioned one-stop threat analysis and monitoring command system can be executed.

[0064] In one embodiment of the present invention, the calculation formula of the first intermediate layer is as follows:

[0065] u (t) =σ(W u X (t) +W u H (t-1) +b u )

[0066] r (t) =σ(W r X (t) +W r H (t-1) +b r )

[0067] C (t) =tanh(W c X (t) +W c r (t) ⊙H (t-1) +b c )

[0068] H (t) =(1-u (t) )⊙C (t) +u (t) ⊙H (t-1)

[0069] Among them, W u 、W r 、W c represents the weight parameter, b u 、b r 、b c represents the bias parameter, ⊙ represents the dot product, u (t) 、r (t) and C (t)Represent the first, second, and third intermediate states respectively, where X (t) Represents the tth data item of one-dimensional structure data, H (t) and H (t-1) They represent the t-th and t-1-th intermediate representation data respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1, H (t-1) =X (t) , tanh is the hyperbolic tangent function, σ represents the sigmoid function;

[0070] The calculation formula for the second intermediate layer is as follows:

[0071]

[0072] where k v The second intermediate representation data of the vth unit of the data matrix No. 1, M (v) is the set of cells associated with the vth cell of the data matrix No. 1, The nth first intermediate representation data output when the first one-dimensional structure data of the uth unit of the first data matrix is ​​input into the first intermediate layer, c v Is a normalization constant, the default value is M (v) The total number of units in , σ is the sigmoid function, W k Represents the third weight parameter.

[0073] The calculation formula of the first output layer is as follows:

[0074]

[0075] Where y represents the first output vector, and its cth component value represents the probability value of the cth scheduling strategy. The scheduling strategy with the largest probability value is selected as the output. The scheduling strategy group contains all the scheduling strategies that can be executed. (all) represents the set of all cells of the data matrix, P1 represents the first image feature, P2 represents the second image feature, W y represents the output weight parameter, b y Represents the output bias parameter.

[0076] It should be noted that the training data of the rail transit operation scheduling model comes from the simulation system. The simulation system simulates the rail transit information that needs to be scheduled and simulates the earthquake source to perform simulation, and obtains training data in the simulation environment.

[0077] The loss function for training the rail transit operation scheduling model is as follows:

[0078]

[0079] SUpeople and SU property Represents the personnel loss and property loss after the scheduling strategy is executed, people max Indicates the total number of passengers on the rail vehicle, property max Indicates the total assets of the rail transit system, WU people and WU property They represent the reduction in personnel losses and property losses compared to not implementing the scheduling strategy.

[0080] The training goal is to minimize casualties and property losses.

[0081] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Ordinary technicians in this field can also make more forms of equivalent embodiments based on the inspiration of this embodiment, all of which are protected by this embodiment.

Claims

1. A method for dispatching urban rail transit operations, characterized in that: The following steps are involved: Step 100: Collecting rail environment information, including earthquake data, population distribution map of the rail operation area, ground road traffic map of the rail operation area, earthquake safety zone map of the rail operation area, rail transportation network topology map, and capacity and status data of all rail vehicles; Step 200: Generate first two-dimensional structural data from the collected track environment information. The first two-dimensional structural data includes a first data matrix and a first relationship matrix. Each cell of the first data matrix represents the first one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transit network topology map, shock-proof safety areas, and earthquake sources. This means that a cell of the first data matrix only contains the track environment information of the independent object it represents. The element in the i-th row and j-th column of the No. 1 relationship matrix represents the association between the i-th unit and the independent object represented by the j-th unit of the No. 1 data matrix. If there is an association, the value of this element in the No. 1 relationship matrix is ​​1, otherwise it is 0; A number one one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the orbital environment information of the corresponding independent object collected at the t-th moment; Step 300: Input the first two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolutional layer, a second convolutional layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolutional layer and the second convolutional layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer. The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers. Step 400: Send the dispatching strategy to the rail transit dispatching center for execution to control the operation of rail vehicles.

2. The urban rail transit operation scheduling method according to claim 1, characterized in that: There is a relationship between rail vehicles if: the rail vehicles are on the same track; There is a relationship between stations if: the stations are located on the same track; The station and the rail vehicle are associated when: the station closest to the rail vehicle is the station; The relationship between the earthquake source rail vehicle and the shock-absorbing safety area means that the rail vehicle and the shock-absorbing safety area are located within the destructive influence range of the earthquake.

3. The urban rail transit operation scheduling method according to claim 1, characterized in that: The orbital environment information of the earthquake source is seismic data.

4. The urban rail transit operation scheduling method according to claim 1, characterized in that: The calculation formula for the first intermediate layer is as follows: u (t) =σ(W u X (t) +W u H (t-1) +b u ) r (t) =σ(W r X (t) +W r H (t-1) +b r ) C (t) =tanh(W c X (t) +W c r (t) ⊙H (t-1) +b c H (t) =(1u (t) )⊙C (t) +u (t) ⊙H (t-1) Among them, W u 、W r 、W c represents the weight parameter, b u 、b r 、b c represents the bias parameter, ⊙ represents the dot product, u (t) 、r (t) and C (t) Represent the first, second, and third intermediate states respectively, where X (t) Represents the tth data item of one-dimensional structure data, H (t) and H (t-1) They represent the t-th and t-1-th intermediate representation data respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1, H (t-1) =X (t) , tanh is the hyperbolic tangent function, and σ represents the sigmoid function.

5. The urban rail transit operation scheduling method according to claim 4, characterized in that: The calculation formula for the second intermediate layer is as follows: where k v The second intermediate representation data of the vth unit of the data matrix No. 1, M (v) is the set of cells associated with the vth cell of the data matrix No. 1, The nth first intermediate representation data output when the first one-dimensional structure data of the uth unit of the first data matrix is ​​input into the first intermediate layer, c v Is a normalization constant, the default value is M (v) The total number of units in , σ is the sigmoid function, W k Represents the third weight parameter.

6. The urban rail transit operation scheduling method according to claim 5, characterized in that: The calculation formula of the first output layer is as follows: Where y represents the first output vector, and its cth component value represents the probability value of the cth scheduling strategy. The scheduling strategy with the largest probability value is selected as the output. The scheduling strategy group contains all the scheduling strategies that can be executed. (all) represents the set of all cells of the data matrix, P1 represents the first image feature, P2 represents the second image feature, W y represents the output weight parameter, b y Represents the output bias parameter.

7. The urban rail transit operation scheduling method according to claim 1, characterized in that: The training data of the rail transit operation scheduling model comes from the simulation system. The simulation system simulates the rail transit information that needs to be scheduled and simulates the earthquake source to perform simulation, and obtains training data in the simulation environment.

8. The urban rail transit operation scheduling method according to claim 7, characterized in that: The loss function for training the rail transit operation scheduling model is as follows: SU people and SU property Represents the personnel loss and property loss after the scheduling strategy is executed, people max Indicates the total number of passengers on the rail vehicle, property max Indicates the total assets of the rail transit system, WU people and WU property They represent the reduction in personnel losses and property losses compared to not implementing the scheduling strategy.

9. An urban rail transit operation and dispatching system, characterized in that: It is used to execute an urban rail transit operation scheduling method according to any one of claims 1 to 8, comprising: The first module collects rail environment information, including earthquake data, population distribution maps of the rail operation area, ground road traffic maps of the rail operation area, earthquake safety zone maps of the rail operation area, rail transportation network topology maps, and capacity and status data of all rail vehicles; The second module generates No. 1 two-dimensional structural data from the collected track environment information. The No. 1 two-dimensional structural data includes No. 1 data matrix and No. 1 relationship matrix. One unit of the No. 1 data matrix represents No. 1 one-dimensional structural data of an independent object. The independent objects include rail vehicles, stations on the rail transportation network topology map, shock-proof safety areas, and earthquake sources. It means that a unit of the No. 1 data matrix only contains the track environment information of the independent object it represents. The track environment information of the earthquake source is earthquake data. The third module inputs the No. 1 two-dimensional structure data, the population distribution map of the rail operation area, and the ground road flow map of the rail operation area into the rail transit operation scheduling model. The rail transit operation scheduling model includes a first convolution layer, a second convolution layer, a first intermediate layer, a second intermediate layer, and a first output layer. The first convolution layer and the second convolution layer respectively input the population distribution map of the rail operation area and the ground road flow map of the rail operation area, and respectively output the first image feature and the second image feature to the first output layer; The first intermediate layer inputs one-dimensional structured data and outputs the first intermediate representation data to the second intermediate layer. The second intermediate layer also inputs the first relationship matrix and outputs the second intermediate representation data to the first output layer. The first output layer outputs the scheduling policy, which includes the next station to which all vehicles need to move, the moving speed, and the operations at the station, including picking up passengers, dropping off passengers, and prohibiting picking up and dropping off passengers. The fourth module sends the scheduling strategy to the rail transit scheduling center for execution to control the operation of rail vehicles.

10. A computer storage medium, characterized in that It is used to store computer-readable instructions, and when the computer-readable instructions are read, it can execute an urban rail transit operation scheduling method as described in any one of claims 1-8.

Images