Train group simulation data generation method based on three-dimension fusion
Through the Gaussian process regression algorithm and the principle of 'seat multiplexing', accurate and comprehensive one-dimensional, two-dimensional and three-dimensional train group simulation data are generated, solving the shortcomings of data generation in the existing technology and ensuring the operation of train group The safety and data integrity provide a valuable research foundation for train group scheduling.
Patent Information
- Application Number
- CN202411110455.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-14
AI Technical Summary
The existing train group simulation operation system lacks comprehensiveness and accuracy when generating simulation data, especially the insufficient utilization of real-time and historical data, and the operation data is not closely related to passenger flow data, making it difficult to generate a comprehensive and accurate train group simulation. data.
The train group simulation data generation method based on three dimensions is adopted. By defining train parameter information and line running scheme, the original operation data is collected, the Gaussian process regression algorithm is used to generate simulation data, and the passenger flow generation is constructed in consideration of the principle of 'seat multiplexing'. The model combines time, line space and passenger flow dimensions to generate simulation data, and makes judgments on the operation safety of train groups, and finally generates a complete train group operation timetable.
It realizes the accurate and comprehensive generation of simulation data of the time axis, line space axis and passenger flow, ensures the safety of train group operation, avoids the generation of unreasonable solutions, and provides a complete train group simulation operation timetable for optimization The train group dispatch plan laid the foundation.
Smart Images

Figure CN119026080A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of transportation network technology, and in particular to a method for generating train group simulation data based on three-dimensional fusion. Background Art
[0002] Although urban rail transit is developing rapidly, the current international situation, industrial revolution, and new urbanization have put forward new requirements and demands for the development of urban rail transit. It is of great research value to scientifically and effectively simulate and optimize the operation and dispatching scheme of trains. In the research and development of train group simulation operation system, it is particularly important to comprehensively and accurately generate the required simulation data, which is the prerequisite for achieving the simulation operation effect.
[0003] At present, the train traction calculation method is often used to generate train simulation operation data, which is mostly the ideal operation results obtained under preset operating conditions. There are also deficiencies in the use of real-time perceived train operation data and historical data. At the same time, most studies are mainly aimed at calculating operation data, and the operation data is not closely related to passenger flow data, and there is a lack of design ideas for comprehensively generating train group simulation data. Summary of the invention
[0004] In order to solve the technical problems existing in the prior art, the present invention provides a train group simulation data generation method based on three-dimensional fusion. Based on the defined train parameter information and line operation plan, the original train operation data is collected and the Gaussian process regression algorithm is used to generate the operation data of the train group simulation. Considering the "seat reuse" principle, a random function is cited to construct a passenger flow generation model to calculate the number of people getting on and off the train group at the station, thereby accurately and comprehensively generating simulation data in three dimensions of time axis, line space axis and passenger flow. At the same time, the particularity of the train group compared with the single train simulation operation is fully considered, and the operation safety of the trains successively dispatched on the same line is judged, and finally a complete train group operation timetable with a reasonable plan is generated.
[0005] The present invention is implemented by the following technical solution: a method for generating train group simulation data based on three-dimensional fusion, comprising the following steps:
[0006] S1. Define the relevant information of the simulated scheme, line, track, station, train and passenger flow, bind the relationship between the defined entities, and set the simulation time step;
[0007] S2, collect the original train operation data, build a Gaussian process regression module to mine the operation rules;
[0008] S3, using the "seat reuse" principle, introduce random functions to build a passenger flow generation module;
[0009] S4, generating simulation data through time and line space dimensions;
[0010] S5. Conduct train group operation safety assessment and operation data integration;
[0011] S6. Supplementing the simulation data of passenger flow dimension at the train arrival time of the train group simulation operation timetable;
[0012] S7. Simulate the operation of the train group according to the three-dimensional data and analyze the operation results of the solution.
[0013] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0014] 1. The present invention is based on the defined train parameter information and line operation plan, uses the Gaussian process regression algorithm to generate the operation data of the train group simulation, considers the "seat reuse" principle to build a passenger flow generation model to calculate the number of people getting on and off the train group at the station, and accurately and comprehensively generates simulation data in three dimensions: time axis, line space axis and passenger flow; and judges the operation safety of trains dispatched on the same line, and generates the data of the subsequent trains after the judgment and storage of the previous train, so as to eliminate unreasonable plans with safety hazards as soon as possible, avoid generating simulation data of the remaining dispatched trains of unreasonable shifts, and cause waste, and finally generate a complete train group simulation operation timetable, which is convenient for subsequent retrieval and simulation operation.
[0015] 2. The present invention generates all the data of the train group simulation operation more accurately and comprehensively, lays the foundation for building a train group simulation operation system, facilitates the research on optimizing the train group dispatching plan, and has good practical application value. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 is a flow chart of the method of the present invention;
[0017] Figure 2 This is an ER diagram of the entity relationships between the various parts of the information defined in this embodiment;
[0018] Figure 3 is the calculation flow chart in step S3 in this embodiment;
[0019] Figure 4 It is a flow chart of train simulation operation data generated by Gaussian process regression;
[0020] Figure 5 It is a flow chart for safety judgment and operation data integration. DETAILED DESCRIPTION
[0021] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.
[0022] Example
[0023] like Figure 1 As shown, in this embodiment, a method for generating train group simulation data based on three-dimensional fusion includes the following steps:
[0024] S1. Define the relevant information of the simulated scheme, line, track, station, train and passenger flow, bind the relationship between the defined entities, and set the simulation time step;
[0025] S2, collect the original train operation data, build a Gaussian process regression module to mine the operation rules;
[0026] S3, using the "seat reuse" principle, introduce random functions to build a passenger flow generation module;
[0027] S4, generating simulation data through time and line space dimensions;
[0028] S5. Conduct train group operation safety assessment and operation data integration;
[0029] S6. Supplementing the simulation data of passenger flow dimension at the train arrival time of the train group simulation operation timetable;
[0030] S7. Simulate the operation of the train group according to the three-dimensional data and analyze the operation results of the solution.
[0031] like Figure 2 As shown, in this embodiment, the definition of relevant information in step S1 is as follows:
[0032] The scheme definition includes: scheme number, number of lines;
[0033] The line definition includes: line number, line name, number of stations;
[0034] Track definition includes: track number, running direction;
[0035] Site definition includes: site number, site name, site longitude and latitude;
[0036] The train definition includes: train number, departure time, maximum running speed, acceleration and deceleration, uniform speed, number of business class seats and first, second and third class seats, train empty seat ratio, and station stop time;
[0037] The definition of passenger flow includes: passenger flow number, passenger flow time period and corresponding percentage of boarding and alighting tickets, and average ticket redemption rate.
[0038] Furthermore, an appropriate simulation time step is set to bind the relationships between entities such as defined schemes and lines, that is, to determine the lines, tracks, stations and trains included in each defined scheme.
[0039] Specifically, in this embodiment, the specific process of step S2 is as follows:
[0040] S21, constructing a Gaussian process regression module, using sensors to collect the running data of the spatial dimension of the train line at each time step and storing it as a training data set of the module, including the train constant speed, maximum speed, starting acceleration, station distance and station travel time;
[0041] S22. Take the constant speed, maximum speed, starting acceleration, station-to-station distance, and station-to-station travel time as input, determine the kernel function and hyperparameters, use the Gaussian process regression algorithm to mine the training data to obtain the operation rules of a single train, and generate the corresponding travel distance of the train at each operating time step during the station-to-station simulation.
[0042] Specifically, in this embodiment, the specific process of step S3 is as follows:
[0043] S31, according to the passenger capacity of each train, the boarding ticket percentage corresponding to the business time period when the train arrives at each station, the alighting percentage and the average ticket redemption rate, the base number of boarding people at the station is calculated as the product of the passenger capacity of the train, the boarding ticket percentage at the station in the time period and the average ticket redemption rate;
[0044] S32. Using the principle of "seat reuse", that is, empty seats can be allocated a corresponding number of additional tickets to the station to accommodate more people, and the newly allocated tickets are converted into additional boarding numbers according to the average ticket exchange rate, and the remaining tickets are allocated to the additional allocated tickets of the next station. Similarly, each station also has tickets that cannot be converted into additional boarding numbers at the previous station, which need to be merged with the additional allocated tickets of the station; the number of boarding people at the station is finally calculated as the sum of the basic boarding number at the station, the additional boarding number and the passenger flow fluctuation range, and the number of alighting people is the product of the number of people on the train and the alighting ratio;
[0045] S33, calculate the number of passengers getting on and off the train at each station by recursion, the calculation process is as follows Figure 3 As shown, i represents the number of trains, j represents the number of time periods, k represents the number of stations, and P ijk 上 , T ijk , Q ijk , R ijk It represents the number of people boarding the i-th train at the k station in the j-time period, the number of people on board, the number of additional boardings, and the number of tickets that could not be converted into additional boardings.
[0046] Specifically, in this embodiment, step S4 generates simulated operation data of each train in the time and line space dimensions by using the Gaussian process regression algorithm; for the simulated data previously generated by the Gaussian process regression algorithm, the simulated data are named and marked with the number of the running train and the names of the two stations between the running stations and stored in the Gaussian process regression operation database. When the simulated data between the same stations of the train needs to be generated again, it is directly called to avoid waste caused by repeated calculations; Figure 4 As shown, the specific process is as follows:
[0047] S41, set the station number n=1, retrieve the train number and the name of the nth and n+1th stations from the database, and for the station to generate simulated operation data, find whether the module operation database has the simulated data previously generated by the train between the stations. If yes, directly retrieve it; if no, find whether the corresponding collected original data is stored in the training database; if there is original data, input the corresponding constant speed, maximum speed, starting acceleration, distance between stations, driving time, and call the Gaussian process regression algorithm to generate simulated operation data between stations and mark it and store it in the operation database; if there is no original data, require the user to provide the original data of the train between the stations, retrain and generate simulated data, and store it in the operation database; with the simulated data, the system calculates the corresponding coordinate data of the train at different time steps between the stations, and then accumulates the running time of the train between the stations in sequence, and combines the train's stay time at the station to obtain the arrival time of the train at the station;
[0048] S42, set n=n+1, and determine whether the station number exceeds the total number of stations N. If so, the loop ends and the generation of the train simulation operation data is completed; otherwise, continue to generate the operation data of the next station of the train, thereby completely generating the train simulation operation data from the time and line space dimensions.
[0049] Specifically, in this embodiment, step S5 uses the maximum running speed v of the train m , train deceleration acceleration a m and safe braking time Calculate the train's emergency braking distance safety distance As the safe distance d between the train and other trains during operation 安 ; According to actual needs, take the minimum relative safe departure time interval t of trains in the same direction 安 , modify this setting as needed; Figure 5 As shown, the specific process of step S5 is as follows:
[0050] S51, sort all defined trains according to departure time, and set the train departure sequence number n=1;
[0051] S52, generating the simulated operation data of the nth departure train according to step S3 and filling it into the train group operation timetable of the scheme;
[0052] S53, setting n=n+1, generating simulated operation data of the nth outgoing train, if the n-1th and nth trains are not running on the same line, directly filling the data of the nth train into the train group operation timetable; if the two trains are bound to the same line, judging the safety of the operation of the two trains;
[0053] S54: Determine whether the terminal station of the nth train is before the starting station of the n-1th train or whether the departure time interval between the two trains is greater than or equal to t 安 , if one of the two is satisfied, the data of the generated nth train will be filled into the planned train group operation schedule; otherwise, from the time when the nth train is dispatched to the time when one of the n-1 and n trains ends its operation, it is determined whether the distance between the two trains at each time step is greater than or equal to the safety distance d of the nth train. 安 If yes, the data of the nth train is filled into the train group operation timetable of the scheme. Otherwise, the scheme is judged to have safety risks and the user is reminded to redefine the simulation information.
[0054] S55. After completing the safety judgment of the nth train, set n=n+1. If n is greater than the total number of trains in the plan, the data of all trains in the plan pass the judgment and are filled in the operation timetable. Otherwise, continue to judge the safety of the nth train until all trains have completed the judgment.
[0055] Specifically, in this embodiment, the specific process of step S6 is as follows:
[0056] S61, input the arrival time of each train at each station obtained in step S4 into the passenger flow generation module in order, and generate the simulation data of the train passenger flow dimension according to step S3, i.e., the number of passengers getting on and off the train at the rear station, according to the percentage of boarding tickets, the percentage of getting off, the average ticket exchange rate and the passenger capacity of the train in the corresponding time period, and add the corresponding consideration data position in the train group simulation operation timetable;
[0057] S62. The system generates all simulation data of the operation plan from three dimensions: time, route space and passenger flow.
[0058] Specifically, in this embodiment, the specific process of step S7 is as follows:
[0059] S71. The departure time of the earliest train is used as the time starting point of the total time axis of the solution, and the time difference between the adjacent simulation data of the train is used as the time iteration interval. Each iteration will cause the trigger function to respond. The trigger function includes the judgment of the departure time of the remaining trains. If the judgment time reaches the departure time of a certain train, a separate corresponding time axis timing trigger is created for the train at the time node, and its starting point is the departure time of the current train; if the judgment result is that the train departure time has not arrived, the time continues to iterate until the last train arrives at the station, the iteration ends, and the time axis is released;
[0060] S72, after the total time axis determines the departure time of a certain train, a time axis corresponding to the current departure train is created. The private time axis takes the departure time of the current train as the starting point and the time when the current train arrives at the last station as the end time. The time iteration interval is consistent with the total axis, and the sub-axis and the total axis iteration are performed synchronously; each iteration system will update the longitude and latitude position, speed, acceleration and other data of the train on each operation chart to determine whether it is the arrival time of the train. If the judgment result is yes, the passenger flow information on the train is updated, otherwise the time continues to iterate; after the train arrives at the last station, the system releases the time axis, thereby realizing the simulation operation of the train group;
[0061] S73, constructing a train operation plan evaluation module, wherein the user sets the evaluation weight, and compares and evaluates the simulation results of various plans based on the average effective seat occupancy rate of the train, the number of passengers sent, and the average number of people carried by the train;
[0062] S74. After the train dispatching plan simulation is completed, the system can quickly retrieve historical data from the database and reproduce the data in the form of various charts. It uses parameter display and graphic display to synchronously present the operating status of different trains in the train group at a certain historical moment, realizing the replay function.
[0063] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.
Claims
1. A method for generating train group simulation data based on three-dimensional fusion, characterized in that: The following steps are involved: S1. Define the relevant information of the simulated scheme, line, track, station, train and passenger flow, bind the relationship between the defined entities, and set the simulation time step; S2, collect the original train operation data, build a Gaussian process regression module to mine the operation rules; S3. Using the "seat reuse" principle, random functions are introduced to build a passenger flow generation module; S4, generating simulation data through time and line space dimensions; S5. Conduct train group operation safety assessment and operation data integration; S6. Supplementing the simulation data of passenger flow dimension at the train arrival time of the train group simulation operation timetable; S7. Simulate the operation of the train group according to the three-dimensional data and analyze the operation results of the solution.
2. A method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: The definition of the relevant information in step S1 is as follows: The scheme definition includes: scheme number, number of lines; The line definition includes: line number, line name, number of stations; Track definition includes: track number, running direction; Site definition includes: site number, site name, site longitude and latitude; The train definition includes: train number, departure time, maximum running speed, acceleration and deceleration, uniform speed, number of business class seats and first, second and third class seats, train empty seat ratio, and station stop time; The definition of passenger flow includes: passenger flow number, passenger flow time period and corresponding percentage of boarding and alighting tickets, and average ticket redemption rate.
3. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: The specific process of step S2 is as follows: S21, constructing a Gaussian process regression module, using sensors to collect the running data of the spatial dimension of the train line at each time step and storing it as a training data set of the module, including the train constant speed, maximum speed, starting acceleration, station distance and station travel time; S22. Take the constant speed, maximum speed, starting acceleration, station-to-station distance, and station-to-station travel time as input, determine the kernel function and hyperparameters, use the Gaussian process regression algorithm to mine the training data to obtain the operation rules of a single train, and generate the corresponding travel distance of the train at each operating time step during the station-to-station simulation.
4. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: The specific process of step S3 is as follows: S31, according to the passenger capacity of each train, the boarding ticket percentage corresponding to the business time period when the train arrives at each station, the alighting percentage and the average ticket redemption rate, the base number of boarding people at the station is calculated as the product of the passenger capacity of the train, the boarding ticket percentage at the station in the time period and the average ticket redemption rate; S32. Utilize the "seat reuse" principle, that is, the corresponding number of tickets can be allocated to the station for empty seats to accommodate more people, and the newly allocated tickets are converted into the number of additional boardings according to the average ticket exchange rate, and the remaining tickets are allocated to the additional allocated tickets of the next station; each station is combined with the additional allocated tickets of the station; the number of boardings at the station is finally calculated as the sum of the baseline number of boardings at the station, the additional number of boardings and the passenger flow fluctuation range, and the number of alightings is the product of the number of boardings and the alighting ratio; S33. Calculate the number of passengers getting on and off the train at each station through recursion.
5. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: Step S4 generates simulated running data of each train in time and line space dimensions by using Gaussian process regression algorithm; the simulated data previously generated by using Gaussian process regression algorithm are named and marked with the number of the running train and the names of the two stations between the running stations and stored in the running database of Gaussian process regression, and directly called when the simulated data between the same stations of the train needs to be generated again; the specific process is as follows: S41, set the station number n=1, retrieve the train number and the name of the nth and n+1th stations from the database, and for the station to generate simulated operation data, find whether the module operation database has the simulated data previously generated by the train between the stations. If yes, directly retrieve it; if no, find whether the corresponding collected original data is stored in the training database; if there is original data, input the corresponding constant speed, maximum speed, starting acceleration, distance between stations, driving time, and call the Gaussian process regression algorithm to generate simulated operation data between stations and mark it and store it in the operation database; if there is no original data, require the user to provide the original data of the train between the stations, retrain and generate simulated data, and store it in the operation database; with the simulated data, the system calculates the corresponding coordinate data of the train at different time steps between the stations, and then accumulates the running time of the train between the stations in sequence, and combines the train's stay time at the station to obtain the arrival time of the train at the station; S42, set n=n+1, determine whether the station number exceeds the total number of stations N, if so, the loop ends, and the train simulation operation data is generated; Otherwise, continue to generate the running data of the train between the next stations, and completely generate the simulated running data of the train from the time and line space dimensions.
6. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: Step S5 uses the maximum running speed v of the train m , train deceleration acceleration m and safety braking time Calculate the train's emergency braking distance safety distance As the safe distance d between the train and other trains during operation 安 ; According to actual needs, take the minimum relative safe departure time interval t of trains in the same direction 安 , modify this setting as needed; the specific process of step S5 is as follows: S51, sort all defined trains according to departure time, and set the train departure sequence number n=1; S52, generating simulation operation data of the nth departure train and filling it into the train group operation timetable of the plan; S53, setting n=n+1, generating simulated operation data of the nth outgoing train, if the n-1th and nth trains are not running on the same line, directly filling the data of the nth train into the train group operation timetable; if the two trains are bound to the same line, judging the safety of the operation of the two trains; S54: Determine whether the terminal station of the nth train is before the starting station of the n-1th train or whether the departure time interval between the two trains is greater than or equal to t 安 If one of the two conditions is satisfied, the data of the nth train generated will be filled into the planned train group operation timetable; Otherwise, from the time when the nth train leaves until one of the n-1 and nth trains finishes running, determine whether the distance between the two trains at each time step is greater than or equal to the safety distance d of the nth train. 安 If yes, the data of the nth train is filled into the train group operation timetable of the scheme. Otherwise, the scheme is judged to have safety risks and the user is reminded to redefine the simulation information. S55. After completing the safety judgment of the nth train, set n=n+1. If n is greater than the total number of trains in the plan, the data of all trains in the plan pass the judgment and are filled in the operation timetable. Otherwise, continue to judge the safety of the nth train until all trains have completed the judgment.
7. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: The specific process of step S6 is as follows: S61, input the obtained arrival time of each train at each station into the passenger flow generation module in sequence, and generate simulation data of the train passenger flow dimension according to the percentage of boarding tickets, the percentage of getting off, the average ticket exchange rate and the passenger capacity of the train in the corresponding time period, that is, the number of passengers getting on and off the train at the subsequent station, and add the corresponding consideration data position in the train group simulation operation timetable; S62. The system generates all simulation data of the operation plan from three dimensions: time, route space and passenger flow.
8. The method for generating train group simulation data based on three-dimensional fusion according to claim 1, characterized in that: The specific process of step S7 is as follows: S71. The departure time of the earliest train is used as the time starting point of the total time axis of the solution, and the time difference between the adjacent simulation data of the train is used as the time iteration interval. Each iteration will cause the trigger function to respond. The trigger function includes the judgment of the departure time of the remaining trains. If the judgment time reaches the departure time of a certain train, a separate corresponding time axis timing trigger is created for the train at the time node, and its starting point is the departure time of the current train; if the judgment result is that the train departure time has not arrived, the time continues to iterate until the last train arrives at the station, the iteration ends, and the time axis is released; S72, after the total time axis determines the departure time of a certain train, a time axis corresponding to the current departure train is created. The private time axis takes the departure time of the current train as the starting point and the time when the current train arrives at the last station as the end time. The time iteration interval is consistent with the total axis, and the sub-axis and the total axis iteration are performed synchronously; each iteration system will update the longitude and latitude position, speed, and acceleration data of the train on each operation chart to determine whether it is the arrival time of the train. If the judgment result is yes, the passenger flow information on the train is updated, otherwise the time continues to iterate; after the train arrives at the last station, the system releases the time axis to realize the simulated operation of the train group; S73, constructing a train operation plan evaluation module, wherein the user sets the evaluation weight, and compares and evaluates the simulation results of various plans based on the average effective seat occupancy rate of the train, the number of passengers sent, and the average number of people carried by the train; S74. After the train dispatching plan simulation is completed, the system retrieves historical data from the database and reproduces the data in the form of various charts. It uses parameter display and graphic display to synchronously present the operating status of different trains in the train group at a certain historical moment, realizing the replay function.
Citation Information
Patent Citations
A method for predicting passenger flow change trend of urban rail transit
CN109272168A
Urban rail transit passenger flow prediction method and system
CN115907181A
Calibration System and Method for Calibrating an Industrial System Model using Simulation Failure
US20230106530A1
Cited By
Passenger flow volume generation method based on ticket amount distribution
CN116703683A
A method for generating passenger flow based on ticket allocation
CN116703683B
