High-speed train operation scheme transportation efficiency evaluation optimization method, device, equipment and storage medium

Abstract

The invention provides a high-speed railway train operation scheme transportation efficiency evaluation optimization method and device, equipment and a storage medium. The method comprises the following steps: acquiring an operation scheme of a high-speed railway train and a passenger load factor of each train; according to the driving scheme of the high-speed railway train and the passenger load factor of each train, determining an optimized driving scheme and a transportation efficiency evaluation value thereof through the evaluation model; wherein the evaluation model comprises a two-dimensional wavelet packet decomposition module and a convolutional neural network; a decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used for converting feature vectors of all trains of a driving scheme into multi-channel data; and the convolutional neural network is used for carrying out transportation efficiency evaluation based on the multi-channel data and the passenger load factor of each train. According to the method, the transportation efficiency is accurately evaluated according to the running scheme of the high-speed railway train and the passenger load factor of each train, the optimized running scheme is obtained, automatic optimization of the transportation efficiency evaluation of the running scheme of the high-speed railway train is realized, and the optimization efficiency and accuracy are improved.

Description

Technical Field

[0001] This application relates to the technical field of rail transit, and particularly to a method, device, equipment, and storage medium for evaluating and optimizing the transportation efficiency of the train operation plan for high-speed trains. Background Art

[0002] In the high-speed railway operation organization planning system, the design of the train operation plan is a key link in balancing supply and demand. The optimization design of the train operation plan not only needs to conform to the travel patterns of passengers and improve the quality of railway transportation services, but also needs to make full use of transport capacity resources and improve the utilization rate of the railway network capacity. Its adjustment and optimization is a complex systematic work.

[0003] In previous studies on the optimization of train operation plans, the relevant optimization methods for train operation plans can be divided into three categories: comprehensive optimization of path design and frequency setting, comprehensive optimization of stop planning and frequency setting, and comprehensive optimization of path design, frequency setting, and stop planning. To evaluate the degree of matching between transport capacity allocation and travel demand, passenger flow distribution will be integrated into the train operation plan optimization process to evaluate the train operation effect. However, due to the complex travel choice behaviors and influencing factors of passenger flow demand, the traditional passenger flow distribution process is difficult to accurately depict the passenger travel path selection and train passenger flow distribution. In the actual transport organization process, usually according to manual experience, the utilization of train transport capacity is estimated based on the advance ticket sales situation of trains.

[0004] This method not only has low efficiency but also lacks accuracy. Summary of the Invention

[0005] To solve one of the above technical defects, this application provides a method, device, equipment, and storage medium for evaluating and optimizing the transportation efficiency of the train operation plan.

[0006] In the first aspect of this application, a method for evaluating and optimizing the transportation efficiency of the train operation plan for high-speed trains is provided. The method includes: Obtain the train operation plan of high-speed railway trains and the occupancy rate of each train; According to the train operation plan of high-speed railway trains and the occupancy rate of each train, determine the optimized train operation plan and its transportation efficiency evaluation value through an evaluation model; Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the train operation plan into multi-channel data; The convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rate of each train; Among them, the convolutional neural network is trained based on multiple historical train operation plans and their train occupancy rates, constructs a mapping relationship, and conducts transportation efficiency evaluation.

[0007] In the second aspect of this application, an optimization device for evaluating the transportation efficiency of high-speed rail train operation plans is provided. The device includes: An acquisition module, configured to acquire the operation plans of high-speed rail trains and the occupancy rates of each train; An optimization module, configured to determine an optimized operation plan and its transportation efficiency evaluation value through an evaluation model according to the operation plans of high-speed rail trains and the occupancy rates of each train; Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data; The convolutional neural network is used to conduct transportation efficiency evaluation based on the multi-channel data and the occupancy rates of each train; Among them, the convolutional neural network is trained based on multiple historical train operation plans and their train occupancy rates, constructs a mapping relationship, and conducts transportation efficiency evaluation.

[0008] In the third aspect of this application, an electronic device is provided, including: A memory; A processor; and A computer program; Among them, the computer program is stored in the memory and is configured to be executed by the processor to implement the method described in the first aspect above.

[0009] In the fourth aspect of this application, a computer-readable storage medium is provided, on which a computer program is stored; the computer program is executed by a processor to implement the method described in the first aspect above.

[0010] The present application provides a method, device, equipment, and medium for evaluating and optimizing the transportation efficiency of a train operation plan. The method includes: obtaining the train operation plan of high-speed railway trains and the occupancy rate of each train; determining the optimized train operation plan and its transportation efficiency evaluation value through an evaluation model according to the train operation plan of high-speed railway trains and the occupancy rate of each train; wherein, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; the decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the train operation plan into multi-channel data; the convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rate of each train; wherein, the convolutional neural network is trained based on multiple historical train operation plans and their train occupancy rates to construct a mapping relationship for transportation efficiency evaluation. The method of the present application can accurately evaluate the transportation efficiency according to the train operation plan of high-speed railway trains and the occupancy rate of each train, and then obtain an optimized train operation plan, realizing the automatic optimization of the transportation efficiency evaluation of the high-speed railway train operation plan, and improving the optimization efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings: Figure 1 is a schematic flowchart of a method for evaluating and optimizing the transportation efficiency of a high-speed railway train operation plan provided by an embodiment of the present application; Figure 2 is a schematic diagram of the principle of an evaluation model provided by an embodiment of the present application; Figure 3 is a schematic diagram of the division of a railway network provided by an embodiment of the present application; Figure 4 is a schematic structural diagram of a device for evaluating and optimizing the transportation efficiency of a high-speed railway train operation plan provided by an embodiment of the present application; Figure 5 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] In order to make the technical solutions and advantages in the embodiments of the present application clearer and more understandable, the following further details the exemplary embodiments of the present application with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than an exhaustive list of all embodiments. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0013] In the process of implementing this application, the inventors found that in previous studies on optimizing train operation plans, the relevant optimization methods for train operation plans can be divided into three categories: comprehensive optimization of route design and frequency setting, comprehensive optimization of stop planning and frequency setting, and comprehensive optimization of route design, frequency setting, and stop planning. To evaluate the matching degree between transport capacity allocation and travel demand, passenger flow distribution is integrated into the process of optimizing the train operation plan to evaluate the effect of train operation. However, due to the complex travel choice behaviors and influencing factors of passenger flow demand, it is difficult for traditional passenger flow distribution processes to accurately depict the travel path selection of passengers and the distribution of train passenger flow. In the actual transport organization process, usually based on manual experience, the utilization of train transport capacity is estimated according to the pre-sale situation of trains. This method not only has low efficiency but also lacks accuracy.

[0014] In view of the above problems, an optimization method, device, equipment, and medium for evaluating the transport efficiency of a train operation plan are provided in an embodiment of this application. The method includes: obtaining the operation plan of high-speed railway trains and the occupancy rate of each train; according to the operation plan of high-speed railway trains and the occupancy rate of each train, determining an optimized operation plan and its transport efficiency evaluation value through an evaluation model; where the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; the decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data; the convolutional neural network is used to evaluate the transport efficiency based on the multi-channel data and the occupancy rate of each train; where the convolutional neural network is trained based on multiple historical operation plans and their train occupancy rates to construct a mapping relationship for transport efficiency evaluation. The method of this application can accurately evaluate the transport efficiency according to the operation plan of high-speed railway trains and the occupancy rate of each train, and then obtain an optimized operation plan, realizing the automatic optimization of the transport efficiency evaluation of high-speed train operation plans and improving the optimization efficiency and accuracy.

[0015] To accurately evaluate the transport efficiency of train operation plans, improve the supply-demand matching degree and the utilization rate of transport capacity resources, an embodiment provides an optimization method for evaluating the transport efficiency of high-speed train operation plans. This method designs a comprehensive multi-dimensional transport efficiency evaluation index system for railway supply-demand efficiency and benefits, and based on historical operation big data, mines the mapping relationship between the characteristics of trains in the train operation plan and transport efficiency to accurately estimate the train operation plan.

[0016] 1. Transport Efficiency Evaluation Index System Since the optimization design of train operation plans needs to consider the benefits of both the railway supply and demand sides, and there is a mutual coordination and restriction relationship between them, therefore, to evaluate the transport efficiency of operation plans, it is necessary to construct the user transport efficiency evaluation index system shown in Table 1 from both the supply and demand aspects.

[0017] Table 1

[0018] (1) The service level evaluation is to evaluate the quality of train services provided for passengers. The evaluation indicators are mainly designed from aspects such as accessibility and convenience, and mainly include the number of passengers arriving and departing, service frequency, O-D coverage, direct transfer ratio, etc., which are used to measure the quality of passengers' travel.

[0019] (2) The operation efficiency evaluation is to evaluate the operation efficiency and capacity utilization of train operations from the perspective of railway operation enterprises, mainly including capacity allocation, running distance, number of car bodies, seat occupancy rate, speed coefficient, operation benefits, etc., which are used to measure the transport organization efficiency of the train operation plan.

[0020] (3) The capacity utilization is to evaluate the utilization rate of resources by the train operation plan from the perspective of railway network transport capacity resources, mainly including the number of trains passing through the section, the number of trains originating and terminating at the station, the section capacity utilization rate, etc.

[0021] To balance the efficiency and benefits of both the supply and demand sides of the railway, the seat occupancy rate of the train is selected as the main indicator for evaluating transport efficiency. The higher the seat occupancy rate of the train, the higher the degree of supply-demand matching and the better the effect of capacity allocation.

[0022] 2. Characteristics of Trains in the Train Operation Plan If the railway network , then the train operation plan .

[0023] Among them, is the set of stations in the railway network, is the set of sections in the railway network.

[0024] is the train identifier, .

[0025] is the total number of stations passed by train , is the set of stations passed by train , , is the first station along the running route of train in the train operation plan, is the second station along the running route of train in the train operation plan, The th station along the running route of train in the train operation plan.

[0026] is the of train The set of stop signs at each station , if the train stops at the first station, then ; otherwise, . If the train stops at the second station, then ; otherwise, . …. If the train stops at the th station, then ; otherwise, .

[0027] is the seating capacity of the train , is the departure time of the train .

[0028] For the characteristics of the trains in the operation plan, including: (1) Running route

[0029] Since the running routes of each train are different, to ensure dimensional consistency, the railway network stations are sorted. If a train passes through a station, it is recorded as 1; otherwise, it is recorded as 0.

[0030] (2) Stop identification

[0031] Similarly, for the sorted railway network stations, if a train stops at a station, it is recorded as 1; otherwise, it is recorded as 0.

[0032] (3) Train seating capacity

[0033] Since the seating capacities of trains of different models vary greatly, for standardization, let the maximum seating capacity among all models be , then the seating capacity of train is converted to , which is the standardized value of .

[0034] (4) Departure time

[0035] For the processing of the departure time, the departure time is converted into the proportion it occupies in the whole day period, that is , which is the standardized value of .

[0036] Then, the characteristic vector of train can be expressed as , and the characteristic vectors of all trains in the operation plan can be expressed as .

[0037] Among them, is the total number of trains in the train operation plan.

[0038] Based on this, referring to Figure 1 , the implementation process of the method for evaluating and optimizing the transportation efficiency of the high-speed rail train operation plan provided in this embodiment is as follows: 101. Obtain the train operation plan of high-speed rail trains and the occupancy rate of each train.

[0039] 102. According to the train operation plan of high-speed rail trains and the occupancy rate of each train, determine the optimized operation plan and its transportation efficiency evaluation value through the evaluation model.

[0040] Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network.

[0041] 1. Two-dimensional wavelet packet decomposition module The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to implement the wavelet packet decomposition process, that is, the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data.

[0042] For example, when the two-dimensional wavelet packet decomposition module implements the wavelet packet decomposition process, it reduces the instability of the original signal by decomposing the original signal into several sub-signals. The principle of wavelet packet decomposition is the spatial decomposition theory of multi-resolution analysis, that is, according to the spatial scale factor, the space is decomposed into several wavelet packet sub-spaces. At the same time, the frequency of the low-frequency sub-space (scale space) and the high-frequency sub-space (detail space) is subdivided, so that the time-frequency resolution of the low and high frequency bands is improved, and better frequency domain localization information than wavelet decomposition is obtained.

[0043] The two-dimensional wavelet packet decomposition module has two decomposition layers, and can convert the feature vectors of all trains in the input operation plan into multi-channel data . The two-dimensional wavelet packet decomposition module selects the function Daubechies4 as the decomposition process function of wavelet packet decomposition, and the generated two-dimensional multi-channel data can be used as the input of the convolutional neural network.

[0044] 2. Convolutional neural network The convolutional neural network is used to evaluate the transportation efficiency based on multi-channel data and the occupancy rate of each train.

[0045] Among them, the convolutional neural network is trained based on historical multiple operation plans and their train occupancy rates, constructs a mapping relationship, and conducts transportation efficiency evaluation.

[0046] A convolutional neural network is a type of feedforward neural network that contains convolutional computations and has a deep structure, capable of learning multi-dimensional data. A convolutional neural network has an input layer, hidden layers, and an output layer. The hidden layers mainly include two types of feature extraction layers, namely, the convolutional layer and the sampling layer (or pooling layer), as well as the fully connected layer.

[0047] Based on the two-dimensional multi-channel data generated by the decomposition of the two-dimensional wavelet packet decomposition module in the evaluation model, the convolutional neural network in the evaluation model has 3 convolutional layers, and each convolutional layer has convolution kernels, and the number of channels is respectively , the activation function is the rectified linear unit; the pooling function uses average pooling, and the last layer of the network is the fully connected layer; the loss function and the optimization function of this convolutional neural network are the mean square error and the Adam algorithm respectively.

[0048] That is, a convolutional neural network includes: an input layer, hidden layers, and an output layer.

[0049] The hidden layers include: a convolutional layer, a pooling layer, and a fully connected layer.

[0050] There are 3 convolutional layers, and each convolutional layer has convolution kernels, and the activation function is the rectified linear unit.

[0051] The pooling function of the pooling layer is the average pooling function.

[0052] The loss function of the convolutional neural network is the mean square error function, and the optimization function is the Adam algorithm.

[0053] In addition, to maximize the transportation efficiency of the train operation plan, the convolutional neural network uses the overall occupancy rate level of the train operation plan as the optimization goal during the transportation efficiency evaluation. Based on the evaluated occupancy rate of the operating trains, the overall occupancy rate level of the operation plan can be calculated as . Since the transportation efficiency is maximization-oriented, to adapt to the minimization-oriented goal of the operation plan optimization model, the train operation plan adjustment and optimization goal are adjusted. Therefore, the objective function during the transportation efficiency evaluation is: .

[0054] Among them, is the train identifier, is the operation plan, is the target value of, is the train the total number of stations passed by, is the train the identifier of the stations passed by, , is the train the seating capacity, is the mileage of the interval , is the interval identifier is the train For the interval of the th station , is the station identifier is the train in the operation plan the th station along the running route of the train is the train in the operation plan the th station along the running route of the train train 's passenger occupancy rate

[0055] The constraint conditions for the pre - evaluation of transport efficiency are: transport capacity allocation constraint, interval passing capacity constraint, number of trains starting and ending at stations constraint, basic diagram constraint, starting and ending time constraint, train seating capacity constraint, service frequency constraint

[0056] 1) Transport capacity allocation constraint The upper limit of railway network transport capacity allocation can be expressed by the train kilometers. According to the formation seating capacity of each train to determine the accounting coefficient of train kilometers : If the train is a single - set train, then ; otherwise . Then the transport capacity allocation constraint is: .

[0057] Among them is the accounting coefficient of the train 's kilometers, is the upper limit of railway network transport capacity allocation (i.e., the upper limit of the available multiple - unit train kilometers of the railway network)

[0058] 2) Interval passing capacity constraint Affected by railway operation safety factors, the number of trains passing through the interval within a unit time (per day or per hour) must meet the corresponding passing capacity limit. Denote the number of trains entering the interval throughout the day as , then the interval passing capacity constraint is: .

[0059] Among them , is the set of intervals of the railway network is the number of trains entering the interval throughout the day is the interval The upper limit of passing capacity.

[0060] 3) Constraints on the number of trains originating and terminating at stations The capacity of trains originating and terminating at stations is affected by various factors such as station grade, nature of technical operations, number of arrival and departure tracks at the station, and distribution of EMU depots (sections). When designing the train operation plan, the number of trains originating and terminating at each station cannot exceed its upper limit of the capacity for originating and terminating trains. The number of originating trains at the station is equal to , and the number of terminating trains is equal to , respectively, and must meet their respective capacity constraints, that is, the constraints on the number of trains originating and terminating at stations are: , .

[0061] Among them, , is the set of stations in the railway network, is the first station along the route of train in the operation plan, is the th station along the route of train in the operation plan, is the upper limit of the capacity for originating and terminating trains at station .

[0062] If station cannot originate and terminate trains, then .

[0063] 4) Basic diagram constraints The trains in the operation plan must come from the alternative trains in the basic operation diagram during the diagram period. Therefore, the basic diagram constraints are: .

[0064] Among them, is the set of trains in the basic operation diagram, .

[0065] 5) Originating and terminating time constraints The originating and terminating times of trains should be within the operation period. Assuming the full-day operation period is , then the originating and terminating time constraints are: .

[0066] Among them, is the starting time of the full-day operation period, is the ending time of the full-day operation period, , is the originating time of train , is the th time of train The arrival time of a station.

[0067] If a given train in the section running time , and at the station the sum of the start-up and stop additional time and the stop time , the arrival time of the train at the stations along the line and the departure time can be determined.

[0068] Among them, is the station identifier, .

[0069] 6) Train capacity constraint The train type selected for the trains in the train operation plan must meet the actually equipped EMU types. Therefore, the train formation must meet the set of feasible capacities of the equipped types, that is, the train capacity constraint is: .

[0070] Among them, is the set of capacities of the equipped types, .

[0071] 7) Service frequency constraint To meet the travel demands of the arriving and departing passengers at the stations, the service at each station must meet a certain service level, that is, there is a certain lower limit of service frequency. Therefore, the frequency of stops at each station must meet this lower limit constraint, that is, the service frequency constraint is: .

[0072] Among them, is the lower limit of the service frequency of station , is the associated identifier of train and station . If train stops at station , then , otherwise .

[0073] When performing step 102, the feature vectors of all trains in the operation plan are a two-dimensional matrix, that is , and the occupancy rate of each train is represented as a one-dimensional label vector, that is . In step 102, refer to Figure 2 ​, based on the eigenvectors of all trains in the train operation plan, they are converted into multi-channel data by using the two-dimensional wavelet packet decomposition process of the two-dimensional wavelet packet decomposition module in the evaluation model; for the decomposed data, the convolutional neural network in the evaluation model is used to perform mapping training between the operation elements of the train operation plan and the passenger occupancy rate of the trains, and then the transportation efficiency of the train operation plan is evaluated to obtain the efficiency evaluation value of the train operation plan. Finally, based on the efficiency evaluation value, the optimized train operation plan and its transportation efficiency evaluation value are determined.

[0074] It should be noted that the above is an explanation of the implementation process of step 102 in terms of principle. The train operation plan here is a superordinate concept. According to the specific situation during execution, the train operation plan can be the train operation plan of high-speed railway trains, or the train operation plan of adjacent trains, or other specific train operation plans. That is to say, the two-dimensional wavelet packet decomposition module in the evaluation model will process the train operation plan, and it does not limit the specific train operation plan to be processed. According to the specific requirements during implementation, an accurate train operation plan can be input into the two-dimensional wavelet packet decomposition module in the evaluation model.

[0075] In specific implementation, the implementation details of step 102 are as follows: 102-1, convert the eigenvectors of all trains in the train operation plan of high-speed railway trains into multi-channel data through the two-dimensional wavelet packet decomposition module .

[0076] Among them, is the train operation plan of high-speed railway trains.

[0077] That is to say, the train operation plan of high-speed railway trains will be input into the two-dimensional wavelet packet decomposition module in the evaluation model, and through the two-dimensional wavelet packet decomposition module, the input eigenvectors of all trains are converted into multi-channel data .

[0078] 102-2, through the convolutional neural network, based on and the passenger occupancy rate of each train, pre-evaluate the transportation efficiency to obtain the efficiency evaluation value . Determine the current solution , determine the objective value of the current solution , determine the optimal solution , determine the objective value of the optimal solution , determine the number of iterations , determine the current temperature .

[0079] Among them, is the preset initial temperature.

[0080] For example, through the convolutional neural network, using the objective function Obtain the performance evaluation value .

[0081] 102-3, according to Construct the operation plan of the neighborhood trains

[0082] In step 102-3, the following processing will be performed on the target trains in to obtain the operation plan of the neighborhood trains: Out-of-service trains Out of service the first target train in

[0083] Among them, the first target train is a single-unit train, and its passenger occupancy rate is lower than , is the preset first ratio value, such as .

[0084] That is to say, for the single-unit trains in the current solution , if the passenger occupancy rate is lower than (such as 25%), then the train will be taken out of service

[0085] Change the train formation capacity 1. Adjust the second target train in to a long formation train with a smaller formation capacity with a probability of .

[0086] Among them, is the preset first probability value, such as . The second target train is a long formation train or a multiple unit train, and its passenger occupancy rate is within , is the preset second ratio value, is the preset third ratio value, , such as , .

[0087] 2. Adjust the third target train in to a single-unit train with a probability of .

[0088] Among them, is the preset second probability value, , such as . The third target train is a long formation train or a multiple unit train, and its passenger occupancy rate is lower than .

[0089] That is to say, for the long formation trains or multiple unit trains in the current solution , if the passenger occupancy rate is within a certain lower range Inside (such as ), then with a certain probability (such as 0.7), it is adjusted to a long formation train with a smaller fixed number of passengers; if the passenger occupancy rate is lower than a certain threshold (such as 40%), then with a certain probability (such as 0.9), it is adjusted to a single - set train.

[0090] Adding trains For long formation / duplex trains with a passenger occupancy rate higher than a certain threshold or short formation trains that cannot perform formation increase, check whether there is a peak line train pair in the adjacent time points in the basic diagram that is similar to the high - utilization section of this train (for an independent peak line, the entire route needs to add trains). If there is, add the corresponding train.

[0091] Therefore, add the train corresponding to the fourth target train in

[0092] Among them, the fourth target train is a long formation train or a duplex train or a short formation train that cannot perform formation increase, and its passenger occupancy rate is higher than , is a preset fourth ratio value, and at the same time, there is a peak line train pair corresponding to the high - utilization section of the fourth target train in the adjacent time points in the basic diagram.

[0093] The train corresponding to the added fourth target train is the train that has a peak line train pair corresponding to the high - utilization section of the fourth target train in the adjacent time points in the basic diagram.

[0094] Adjusting stops For the station , if it is to meet the lower limit constraint of service frequency, taking the principle that the stop distribution of the train at the station satisfies space - time uniformity, select a train with a lower stop - through rate to add stops without causing operation conflicts, and update the arrival and departure times of the train along the line.

[0095] Therefore, adjust the arrival and departure times along the line of the fifth target train in

[0096] Among them, the stations along the line of the fifth target train include the target station, and the target station meets the service frequency constraint.

[0097] The service frequency constraint is: , where is the train identification, is the operation plan, is the lower limit of the service frequency of the station , is the train and the station 's association identification, the train At the station If it stops at the station otherwise .

[0098] Adjust the operating section For trains with a low passenger occupancy rate, determine the start and end points of the extended section based on the capacity constraints of the originating and terminating stations and the section adjacent to the origin and destination, and adjust the operating section of the train.

[0099] Therefore, adjust according to the capacity constraint of the section the operating section of the sixth target train.

[0100] Among them, the passenger occupancy rates of the sixth target trains are all lower than , which is the preset fifth proportional value.

[0101] The capacity constraint of the section is as follows: .

[0102] Among them, is the section identifier, , is the set of sections of the railway network, is the number of trains entering the section throughout the day, is the section the upper limit of the passing capacity.

[0103] It should be noted that this embodiment does not limit the relationship between the above , , , . Their values can be exactly the same, partially the same, or completely different. Similarly, this embodiment does not limit the relationship between the above , , , . Their values can be exactly the same, partially the same, or completely different. But it needs to satisfy .

[0104] 102 - 4. Through the two-dimensional wavelet packet decomposition module, the eigenvectors of all trains in the operation plan of the neighboring trains are converted into multi-channel data .

[0105] Among them, is the operation plan of the neighboring trains constructed in the th iteration.

[0106] That is to say, the operation plan of the neighboring trains will be In the input evaluation model, the two-dimensional wavelet packet decomposition module converts the eigenvectors of all trains in into multi-channel data through the two-dimensional wavelet packet decomposition module. .

[0107] 102-5. Based on and the occupancy rate of each train, pre-evaluate the transport efficiency through a convolutional neural network to obtain the efficiency evaluation value . .

[0108] For example, through a convolutional neural network, use the objective function to obtain the efficiency evaluation value .

[0109] 102-6. Determine the evaluation difference .

[0110] 102-7. If , then update , update , update , update .

[0111] If but , then update , update .

[0112] Among them, is the natural exponential function, and is the random function.

[0113] 102-8. .

[0114] 102-9. If it is determined that the termination condition is not satisfied according to and , then re-execute the steps of constructing the operation plan of the neighboring trains according to (i.e., step 102-3) and the subsequent steps for cycling. If it is determined that the termination condition is satisfied according to and , then determine that the optimized operation plan is , and the transport efficiency evaluation value of the optimized operation plan is .

[0115] Specifically, if , then determine that the termination condition is not satisfied.

[0116] Among them, is the preset total number of iterations.

[0117] If , then update ,renew .like , then it is determined that the termination condition is not met. If , then it is determined that the termination condition is met.

[0118] in, is the temperature drop ratio, is the preset end temperature.

[0119] In the specific implementation, step 102 will be based on the railway network , all-day operation , the operation plan of high-speed railway trains , the train set of the basic operation diagram , the upper limit of railway network capacity configuration ,station The starting and ending capacity limit , Train at the station The sum of the start-stop additional time and the stop time , Service frequency lower limit , interval The upper limit of the passing capacity ,train Calculation factor for kilometers , initial temperature , termination temperature , the total number of iterations , temperature drop ratio , the passenger load factor of each train, and the optimized operation plan is obtained through the evaluation model and its transport efficiency evaluation value

[0120] Train formation calculation coefficient , algorithm initial temperature and termination temperature , the number of inner loop iterations , temperature drop ratio .

[0121] In addition, the high-speed railway train operation plan in the high-speed railway train operation plan transportation efficiency evaluation optimization method provided in this embodiment is pre-compiled, and the high-speed railway train operation plan transportation efficiency evaluation optimization method provided in this embodiment can directly obtain the compiled high-speed railway train operation plan. The compilation process of the high-speed railway train operation plan is as follows: 201, determine the complete set of passenger train operation plans for the specified lines or areas in the next chart period.

[0122] When performing step 201, the full set of passenger train operation plans for a specified line or area in the next diagram period can be determined based on the pre-input estimated data of passenger travel demands, combined with the passenger train operation plan of the current diagram period, and comprehensively considering the new plan, cancellation plan, and adjustment plan.

[0123] In addition, the core of formulating the operation plan of high-speed railway trains lies in how to determine whether the set of operation plans in the current state meets the constraint conditions of the HHTPV (High-feasibility High-speed Train Plan Validation) framework. Among them, the passenger trains corresponding to the set of operation plans in the current state can be divided into two categories. One category is the passenger trains that have been clearly determined to operate and have been drawn on the operation diagram. The set of such operation plans is called the set of operation plans for regularly operating passenger trains. For newly built high-speed railway lines, such operation plans mainly refer to the operation plans that need to be implemented on the high-speed railway line, such as the benchmark trains operating on this line, or the important trains starting or ending at key hub stations and passing through this line. Although such trains have not been drawn on the operation diagram at the initial stage of formulating the operation plan of high-speed railway trains, due to the importance of relevant plans, they need to be determined to be implemented and drawn at the initial stage of formulating the operation plan of high-speed railway trains. For non-newly built high-speed railway lines, in addition to the above situations, such operation plans mainly include the operation plans of passenger trains that have been drawn on the operation diagram and have not been cancelled in the current diagram period.

[0124] The other category is the set of alternative passenger train operation plans. At the initial stage of formulating the operation plan of high-speed railway trains, such operation plans have not been fully drawn on the operation diagram, and the number of operation plans can be greater than the number of trains that actually need to operate and the carrying capacity of the railway network. By setting a redundant set of alternative operation plans, the result of the iterative process of the formulation process can be made more flexible. At the same time, during the formulation process, the alternative train operation plans can be adjusted according to the actual situation to obtain a better operation plan.

[0125] It should be noted that in order to more accurately implement the optimization process of operation plans, a certain degree of redundancy setting is carried out on the operation plans when initializing the set of alternative operation plans. The purpose of the redundancy setting is to finely adjust the alternative train operation plans according to the actual situation during the formulation process to obtain a better comprehensive operation plan. In practical applications, the number of operation plans with redundancy setting should be reasonably set according to the actual situation to ensure the flexibility and efficiency of the formulation process.

[0126] In summary, the full set of passenger train operation plans for a specified line or area in the next diagram period determined in step 201 .

[0127] Among them, is the set of passenger train operation plans that are fixed for specified lines or regions.

[0128] , is the set of trains that meet the preset conditions in the passenger train operation plans newly planned for specified lines or regions, is the set of passenger train operation plans that have been drawn in the train operation diagram for specified lines or regions and have not been cancelled during the current diagram period.

[0129] , is the set of passenger train operation plans for specified lines or regions during the current diagram period, is the set of passenger train operation plans for specified lines or regions that are planned to be out of service.

[0130] is the set of alternative passenger train operation plans for specified lines or regions in the next diagram period, , is the set of passenger train operation plans newly planned.

[0131] It should be noted that in the operation plan sets of this embodiment and subsequent embodiments, it is allowed that there are cases where the basic operation information of train plans is the same but the serial numbers are different.

[0132] 202. Based on the basic data of the railway network, determine the basic constraints of the HHTPV framework.

[0133] When performing step 202, based on the specified lines or regions targeted by the operation plan compilation, according to the pre-input basic data of the railway network, comprehensively consider the topological connectivity of the railway network, deduce the passing capacities of key intervals and stations (yards), calculate the basic parameters of the HHTPV framework, and determine the basic constraints of the HHTPV framework. This basic constraint is the static constraint of the HHTPV framework.

[0134] Therefore, in the specific implementation, the implementation process of step 202 is as follows: 202-1. Based on the basic data of the railway network and the topological connectivity of the railway network, deduce the upper limit of the passing capacity of intervals, the upper limit of the passing capacity of stations, the stop capacity limit value of stations, and the limited quantity of each train type within the interval.

[0135] 202-2. According to the upper limit of the passing capacity of intervals, the upper limit of the passing capacity of stations, the stop capacity limit value of stations, and the limited quantity of each train type within the interval, determine the passing limit constraints and train type limit constraints of each interval, and the passing limit constraints and stop limit constraints of each station.

[0136] 1. Passing limit constraints of intervals Among them, the passing restriction of any interval is , is the interval identifier, is the train identifier, is the complete set of passenger train operation plans for the specified line or area in the next diagram period, is the train in the interval the marked parameter of passing, is the train activation parameter, is the interval upper limit of passing capacity.

[0137] is 0 or 1, indicating that the train passes through the interval , indicating that the train does not pass through the interval .

[0138] is 0 or 1, indicating that the train is not activated, indicating that the train is activated.

[0139] 2. Train type restriction constraint of the interval The train type restriction constraint of any interval is , is the train type identifier, is the train belonging to the train type marked parameter, interval the limit of train type within.

[0140] is the train belonging to train type.

[0141] Subject to the comprehensive influence of operation plan and passing capacity limit.

[0142] 3. Passing restriction constraint of the station The passing restriction constraint of any station is , is the station identifier, is the train at the station the marked parameter of passing, is the station The upper limit of passing capacity.

[0143] Is 0 or 1, Indicates that the train At the station Passes by, Indicates that the train Not at the station Passes by.

[0144] 4. Stopping restrictions and constraints at stations The stopping restriction and constraint for any station is , For the train At the station The marking parameter for stopping, For the station The stopping capacity limit value.

[0145] Is 0 or 1, Indicates that the train At the station Stops, Indicates that the train Not at the station Stops.

[0146] Steps 201 and 202 are the preprocessing stage for formulating the train operation plan of high-speed railway trains. Through this stage, the determination of the basic constraints of the HHTPV framework and the generation of the complete set of passenger train operation plans for the specified line or area in the next diagram period are carried out.

[0147] 203, Initialize the HHTPV framework based on the constraint conditions.

[0148] Among them, the constraint conditions are determined according to the basic constraints and the dynamic threshold constraint set.

[0149] Step 203 is the initialization process of the HHTPV framework. In step 203, the HHTPV framework will be initialized according to the constraint conditions of the HHTPV framework in the current state, and the preparatory work of the HHTPV framework will be completed.

[0150] Among them, the constraint conditions include the basic constraints obtained in step 202 and the dynamic threshold constraint set obtained in step 205.

[0151] In step 205, an iterative process of the operation diagram coordination stage will be carried out. This iterative process will obtain the latest train operation constraints, and this latest constraint constitutes an ever-updating dynamic threshold constraint set.

[0152] 204. Generate the train operation plan in the current state according to the preferred strategy, the preferred adjustment direction of the train operation plan, and the complete set of train operation plans for the specified line or area in the next diagram period.

[0153] Step 204 is the process of optimizing the train operation plan. In this process, according to the pre-input preferred strategy and combined with the preferred adjustment direction of the train operation plan (the preferred adjustment direction of the train operation plan is the adjustment feedback information obtained through iterative feedback in the process of plan compilation), optimize the alternative train operation plans in it to generate the train operation plan in the current state.

[0154] In addition, this embodiment does not limit the preferred strategy, which can be determined according to actual needs. For example, if the actual need is passenger demand (such as the fewest train stops), then under the constraint of meeting passenger demand, with the goal of the fewest train stops, determine the preferred strategy as .

[0155] Among them, is the train identification, is the complete set of train operation plans for the specified line or area in the next diagram period, is the station identification, is the set of stations of the railway network, is the train at the station stop marking parameter.

[0156] In addition to the above preferred strategy, a weighted multi-objective preferred strategy can also be proposed according to the optimization objectives of multiple types of train operation plans. For example, simultaneously set objectives such as the fewest train stops, uniform stop distribution, and reasonable train type setting as multi-objectives for weighting to form a multi-objective preferred strategy. In reality, the objectives of the preferred strategy have a certain degree of complexity, and some objectives cannot be numerically quantified. In this case, according to the specific situation, an artificial preferred strategy can be adopted, and professional personnel can optimize the train operation plan according to the actual situation to ensure the rationality of the train operation plan.

[0157] Steps 203 and 204 are the plan compilation stage of the train operation plan compilation for high-speed railway trains. Through this stage, initialize the HHTPV framework, execute the iteration of the train operation plan optimization process, gradually optimize the train operation plan for passenger trains, and finally obtain the train operation plan in the current HHTPV framework state.

[0158] 205. In the HHTPV framework, verify the set of train operation plans in the current state, and compile the train operation plan for high-speed railway trains according to the verification results.

[0159] Among them, the verification includes constraint verification and feasibility verification. Therefore, the implementation process of step 205 is as follows: within the HHTPV framework, perform constraint verification on the set of train operation plans in the current state.

[0160] If the constraint verification fails, then: 1) Update the preferred adjustment direction of the train operation plan.

[0161] 2) Re-execute the steps of generating the train operation plan in the current state according to the preferred strategy, the preferred adjustment direction of the train operation plan, and the complete set of train operation plans for the specified line or area in the next time period (i.e., step 204) and subsequent steps.

[0162] If the constraint verification passes, then: 1) Perform preliminary drawing of the train operation diagram on the set of train operation plans in the current state to generate the result of preliminary drawing of the train operation diagram.

[0163] In the preliminary drawing of the train operation diagram, the preliminary drawing of the train operation diagram can be performed according to the set of train operation plans in the current state generated in the plan compilation stage to generate the result of preliminary drawing of the train operation diagram. To balance the accuracy and efficiency of the preliminary drawing of the train operation diagram, a train operation constraint model can be established, and the preliminary drawing of the train operation diagram can be automatically completed through a computer program to improve the efficiency of the entire process. In specific implementation, since the preliminary drawing process is to verify the high feasibility of the train operation plan, some of the constraints for the actual drawing of the train operation diagram can be appropriately relaxed, reducing the difficulty of drawing the train operation diagram, reducing resource consumption in the verification process, and improving the verification efficiency.

[0164] 2) Perform feasibility verification based on the result of preliminary drawing of the train operation diagram.

[0165] The feasibility verification process can be realized based on the discrimination of train operation diagram conflicts. For example, based on the result of preliminary drawing of the train operation diagram, determine whether the set of train operation plans in the current state is feasible.

[0166] 3) If the feasibility verification fails, update the set of dynamic threshold constraints, and re-execute the steps of initializing the HHTPV framework based on the constraint conditions (i.e., step 203) and subsequent steps.

[0167] If the feasibility verification passes, determine the set of train operation plans in the current state as the train operation plan for the high-speed railway trains to be compiled.

[0168] For example, if the feasibility verification fails, through the analysis of train operation diagram conflicts, generate new dynamic threshold constraints, form a new set of dynamic threshold constraints, introduce them into the HHTPV framework, and re-execute the process of the plan compilation stage until the final train operation plan with high feasibility is generated. This final train operation plan with high feasibility is the train operation plan for the high-speed railway trains to be compiled.

[0169] If the feasibility verification is passed, there is no train operation conflict. At this time, the set of train operation plans in the current state is the passenger train operation plan with high feasibility. This passenger train operation plan with high feasibility is the final train operation plan for high-speed railways compiled. This train operation plan for high-speed railways meets the constraint conditions of the HHTPV framework, has high feasibility, and can enter the subsequent railway passenger transport operation organization process to provide efficient and safe travel services for passengers.

[0170] Among them, the process of updating the dynamic threshold constraint set is as follows: perform operation diagram conflict analysis based on the pre-drawn results of the operation diagram, and update the dynamic threshold constraint set through Update the dynamic threshold constraint set.

[0171] Among them, is the train identification, is the complete set of passenger train operation plans for the specified line or area in the next diagram period, is the train type identification, is the set of train types, is the section identification, is the set of sections of the railway network, is the dynamic threshold constraint identification in the dynamic threshold constraint set, is the dynamic threshold constraint belongs to the section marking parameter, is for the train in the section passing through marking parameter, is for the train belonging to the train type marking parameter, is the dynamic threshold constraint constraint weighting for the train type , is the enabling parameter for the train , is the dynamic threshold constraint limit value.

[0172] is 0 or 1, indicating that the dynamic threshold constraint belongs to the section , indicating that the dynamic threshold constraint does not belong to the section. It is allowed to manage constraints for multiple related sections, so as to achieve the resolution of train conflicts in multi-section management and expand the scope of action of the constraints.

[0173] , that is, the section is composed of partially ordered station pairs.

[0174] is a set of train types, that is, a refined set of train categories, where subdivision is allowed according to train grade, train speed, and whether it belongs to a benchmark train.

[0175] With the dynamic threshold constraint limiting value cooperation, the control ability for grouping multiple types of trains can be achieved.

[0176] The iterative update process of the dynamic threshold constraint set in this step is an important part of the HHTPV framework. Its core role is to verify the high feasibility of the train operation plan. Through the multi-group dynamic control of the carrying capacity of the train operation plan in the line section, if combined with the operation plan optimization process in step 204, a mathematical programming problem for optimizing the passenger train operation plan can be constructed inside the closed-loop iterative process in step 205 and calculated and solved. At the same time, verification is carried out through dynamic threshold constraints to ensure that the train operation plan of high-speed railway trains conforms to the limitations of the current constraint combination.

[0177] The dynamic threshold constraint generation process adopts a closed-loop iterative mechanism. First, based on the current operation plan, the train operation diagram is pre-paved. The section with irreconcilable conflicts is identified through the conflict detection algorithm. Subsequently, threshold dynamic adjustment is performed for the problem section: modifying the existing threshold constraint limiting value parameters, adding targeted grouping threshold constraints to exclude conflicting operation plan combinations, and cleaning up the redundant threshold constraints generated in the historical iteration. After multiple rounds of cyclic iteration of "plan optimization - conflict detection - threshold constraint adjustment", a high-feasibility operation plan that not only meets the transportation capacity limitations but also conforms to the optimization mechanism is finally generated, that is, the train operation plan of high-speed railway trains.

[0178] This constraint condition has different application characteristics in different scenarios. For non-core lines or regions, the feasibility judgment can be quickly completed by simplifying the threshold parameter settings of relevant sections, improving the efficiency of plan compilation; while in core lines or regions, the grouping constraints can be refined based on dimensions such as train speed and vehicle formation, and the combination characteristics of the train operation plan can be accurately controlled through multi-level threshold constraints to ensure the high feasibility of the train operation plan of high-speed railway trains.

[0179] After the dynamic threshold constraint set initializes the HHTPV framework in step 203, the dynamic threshold constraint set remains unchanged in step 204 (that is, the dynamic threshold constraint set is fixed in the operation plan optimization process in step 204), but in the iterative process of step 205, it will be dynamically adjusted according to the result of the pre-paved train operation diagram.

[0180] In addition, by adjusting such that Equal to 1, the parameters can be reduced, which corresponds to the new constraint . However, keeping can make the constraint more in line with the physical meaning of train operation. In practical applications, and should be reasonably set according to the characteristics of the train type to ensure the effectiveness of the constraint.

[0181] Step 205 is the operation diagram coordination stage for the preparation of the operation plan of high-speed railway trains. Through this stage, the operation plan set in the current state can be verified according to the constraint conditions of the HHTPV framework in the current state to determine whether it meets the constraint conditions of the HHTPV framework. If it is satisfied, it enters the pre-drawing stage of the operation diagram; if it is not satisfied, a new preferred adjustment direction for the operation plan is generated according to the adjustment feedback information of the HHTPV framework. Steps 204 and subsequent steps are re-executed until the constraint conditions of the HHTPV framework are met.

[0182] In the pre-drawing stage of the operation diagram, the operation diagram of the operation plan set in the current state can be pre-drawn to determine whether it is feasible. For the infeasible operation plan, a new set of dynamic threshold constraints is generated through the operation diagram conflict discrimination, and the HHTPV framework is introduced, and steps 203 and subsequent steps are re-executed until the final highly feasible operation plan of passenger trains is generated. The final highly feasible operation plan of passenger trains is the operation plan of high-speed railway trains prepared.

[0183] The process of preparing the operation plan of high-speed railway trains provided in this embodiment can be applied to any railway network structure. However, for the execution efficiency of the preparation of the operation plan of high-speed railway trains and the accuracy of the results, the railway network can be reasonably divided and transformed so as to verify the operation plan involving the specified line or area in a targeted manner. Although the railway network division reduces the verification scope of the HHTPV framework, through the combination of the static constraints and dynamic threshold constraints of the HHTPV framework, the feasibility of the operation plan of trains outside the verification scope can still be evaluated.

[0184] The division of the railway network should be based on the actual situation to ensure that the divided railway network structure has a certain degree of independence while being able to interact effectively with other railway network structures. When dividing the railway network, the following factors can be considered. First, the core lines or regions for the compilation of the train operation plan need to be completely retained to ensure the complete and accurate feasibility verification of the train operation plan in the core lines or regions. Second, for the lines connected to the core lines or regions but without cross-line operation in the train operation plan, since they are not involved in the high-feasibility verification of the train operation plan, they can be directly deleted from the railway network to reduce the computational workload. Finally, for the lines connected to the core lines or regions and with cross-line operation in the train operation plan, it is necessary to conduct a simplified high-feasibility verification of the relevant train operation plans according to the carrying capacity of the relevant lines and the actual train operation conditions. Therefore, the basic connection relationship with the core lines or regions needs to be retained, that is, the intermediate stations on other lines in the route are deleted, and the connection relationship with the core lines or regions is established through the way of virtual sections to ensure the high feasibility of the relevant train operation plans.

[0185] Figure 3 Fig. shows a schematic diagram of the division of a railway network. Figure 3 The "road network" in Fig. refers to the railway network. Figure 3 Fig. shows the core section of line A composed of stations A1 to A7 for the compilation of the train operation plan, and simplifies the complete railway network (i.e., Figure 3 the complete road network in Fig.) to generate a converted railway network (i.e., Figure 3The result of the converted network). Lines B, C, D, and E represent four lines connected to Line A respectively. Through the connection of Line A with other lines, the interaction between Line A and other lines can be achieved. For Line B, assuming that there is no existing or upcoming train operation plan, and crossing from Station A4 to Station B1, then Line B in the converted network can be deleted. For Line C, assuming there is a train operation plan for crossing lines, crossing from Station A5 to Stations C1 and C3, and the final relevant operation plan has C5 as the origin and destination station, then Line C in the converted network is retained, where Stations C1 and C5 are retained to completely model the relevant operation plan. Since Line C is not the core line for the preparation of this operation plan, in the converted network, the structure of Line C is simplified, and only the connection relationship with Line A is retained to reduce the calculation amount. At the same time, in the operation process of the HHTPV framework, through the dynamic threshold constraints related to the section from Station C1 to Station A7 and the virtual section from Station A7 to Station C5, the high feasibility of the train operation plan for crossing to Line C is ensured. For Lines D and E, assuming there is a relevant operation plan passing through Station D3 and having Station E1 or Station E5 as the origin and destination station, then Lines D and E in the converted network are retained, where Stations D3, E1, and E5 are retained to completely model the relevant operation plan. In the operation process of the HHTPV framework, by establishing the dynamic threshold constraints of the corresponding virtual section, the high feasibility of the relevant train operation plan is ensured.

[0186] The process of formulating the train operation plan for high-speed railway trains is divided into three stages. After the preprocessing stage, the latter two stages are iteratively looped. That is, in the preprocessing stage, the estimated passenger demand data is input to determine the complete set of passenger train operation plans, and the basic data of the railway network is input to determine the basic constraints of the HHTPV framework; in the plan formulation stage, enter the iteration, initialize the HHTPV framework, repeatedly optimize the passenger train operation plan, and obtain the operation plan in the current state; in the train operation diagram coordination stage, execute the pre-drawing process of the train operation diagram, judge whether the operation plan generated in the plan formulation stage can be drawn in the train operation diagram. If it is not feasible, through the discrimination of train operation diagram conflicts, calculate the stations or sections where conflicts may occur, generate a new combination of dynamic threshold constraints for high-speed rail trains according to the corresponding topological structure, and introduce it into the HHTPV framework. Otherwise, complete the formulation of the final operation plan to obtain the train operation plan for high-speed railway trains.

[0187] The process of formulating the train operation plan for high-speed railway trains provided in this embodiment is based on the basic data of the railway network to determine the basic constraints of the HHTPV framework; initialize the HHTPV framework based on the basic constraints, verify the set of operation plans in the current state within the HHTPV framework, and determine the formulated operation plan according to the verification results, realizing the automatic verification of the operation plan and improving the formulation efficiency.

[0188] In addition, on the basis of reasonable verification in the preparation of the operation plan, the generation of the railway passenger transport timetable can be carried out, which is equivalent to the optimization of the whole chain. The main considerations for the generation of the railway passenger transport timetable are as follows: The passenger flow information of the current station. Through the passenger flow information of the current station, the passenger flow heat is learned to obtain the passenger flow heat of the current station, the local area where the current station is located is determined, and the local heat matrix is constructed according to the passenger flow heat of other stations corresponding to the local area where the current station is located.

[0189] Perform eigenvalue decomposition on the local heat matrix to obtain the heat learning features corresponding to the local area where the current station is located, and generate the initial passenger transport timetable of the current station through the heat learning features. And determine the in-station dispatching coupling degree of the current station through the passenger flow heat and the passenger transport adjustment ratio of the current station.

[0190] Determining the in-station dispatching coupling degree of the current station specifically includes: Based on the preset coupling degree extraction period, trend values of the passenger flow heat change trend and the passenger transport adjustment trend are respectively collected to obtain a passenger flow heat trend value sequence and a passenger transport adjustment trend value sequence; After normalizing the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence, perform coupling degree analysis according to the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence to obtain the in-station dispatching coupling degree of the current station.

[0191] In specific implementation, the coupling degree extraction period can be obtained by mapping based on the current passenger flow density. In some embodiments, the coupling degree extraction period can be fixed as a constant value of 6h, and then the Pearson correlation coefficient between the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence can be used as the in-station dispatching coupling degree.

[0192] Dynamically feedback and update the initial passenger transport timetable according to the in-station dispatching coupling degree and the inter-station dispatching coupling degree to generate an optimized passenger transport timetable.

[0193] Aiming at the problems of insufficient dynamic response, difficult multi-professional collaboration and low calculation efficiency caused by the independent operation of the passenger flow prediction, train adjustment, seat allocation and timetable compilation links in the traditional method, a multi-professional modular collaborative system for passenger transport, transportation, dispatching and vehicles is constructed, a closed-loop process of "adjustment plan generation → timetable dynamic generation → operation evaluation → feedback optimization" is established, and lightweight data interaction is realized by using a standardized interface.

[0194] The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment starts from the efficiency and benefits of both the railway supply and demand sides, selects the overall passenger occupancy rate level of the operation plan to construct a transportation efficiency evaluation index system, and can evaluate the transportation efficiency of the operation plan from multiple dimensions such as service level, operation efficiency, and capacity utilization.

[0195] The evaluation model designs a two-dimensional feature matrix based on the operation elements of the operation plan, performs two-dimensional wavelet packet decomposition using the two-dimensional wavelet packet decomposition module, converts it into multi-channel data, and uses a convolutional neural network to evaluate the transportation efficiency.

[0196] When evaluating the transportation efficiency, taking the overall passenger occupancy rate level as the optimization goal of the operation plan, comprehensively considering constraint conditions such as transport capacity allocation, transportation organization, and service level, construct the operation plan of neighboring trains according to the transportation efficiency evaluation results, and solve it.

[0197] The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment can improve the accuracy of the transportation efficiency evaluation of the operation plan. The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment comprehensively extracts the characteristic elements of the operation plan based on historical operation big data, deeply excavates the mapping and correlation relationships between the operation elements of the train operation plan and between the operation elements and the transportation efficiency, realizes the prediction of the passenger occupancy rate distribution of the operating trains, and improves the accuracy of the transportation efficiency evaluation of the operation plan.

[0198] The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment can improve the supply-demand adaptability of the operation plan. The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment uses the pre-evaluation technology of the transportation efficiency of the operation plan to obtain the distribution of the supply-demand adaptability of the operation plan, designs corresponding adjustment strategies for the operation elements according to the distribution of the supply-demand adaptability, and uses the iterative algorithm to continuously optimize the operation plan to improve the supply-demand adaptability of the operation plan.

[0199] The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment can optimize the transportation organization efficiency. The method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan provided in this embodiment constructs a pre-evaluation model for the transportation efficiency of the operation plan, accurately and quantitatively evaluates the transportation efficiency of the operation plan, provides a reference basis for the optimization design of the operation plan, and at the same time designs an optimization method for the operation plan based on this, realizes the intelligent decision-making of the train operation plan, optimizes the implementation effect of the train operation, effectively reduces the manual workload, and optimizes the transportation organization efficiency.

[0200] This embodiment provides a method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan, which obtains the operation plan of high-speed railway trains and the occupancy rate of each train; according to the operation plan of high-speed railway trains and the occupancy rate of each train, the optimized operation plan and its transportation efficiency evaluation value are determined through an evaluation model; among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; the decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data; the convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rate of each train; among them, the convolutional neural network is trained based on multiple historical operation plans and the occupancy rate of their trains, constructs a mapping relationship, and conducts transportation efficiency evaluation. The method of this embodiment can accurately evaluate the transportation efficiency according to the operation plan of high-speed railway trains and the occupancy rate of each train, and then obtain an optimized operation plan, realizing the automatic optimization of the transportation efficiency evaluation of the high-speed train operation plan, and improving the optimization efficiency and accuracy.

[0201] Based on the same inventive concept of the method for evaluating and optimizing the transportation efficiency of the high-speed train operation plan, this embodiment provides a device for evaluating and optimizing the transportation efficiency of the high-speed train operation plan. Refer to Figure 4 , the device includes: An acquisition module 401, configured to acquire the operation plan of high-speed railway trains and the occupancy rate of each train.

[0202] An optimization module 402, configured to determine an optimized operation plan and its transportation efficiency evaluation value through an evaluation model according to the operation plan of high-speed railway trains and the occupancy rate of each train acquired by the acquisition module 401.

[0203] Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network.

[0204] The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data.

[0205] The convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rate of each train.

[0206] Among them, the convolutional neural network is trained based on multiple historical operation plans and the occupancy rate of their trains, constructs a mapping relationship, and conducts transportation efficiency evaluation.

[0207] Among them, the optimization module 402 is configured to convert the feature vectors of all trains in the operation plan of high-speed railway trains into multi-channel data through the two-dimensional wavelet packet decomposition module . Among them, is the operation plan of high-speed railway trains.

[0208] Based on the convolutional neural network and the occupancy rate of each train, perform a pre - evaluation of the transportation efficiency to obtain the efficiency evaluation value . Determine the current solution , determine the objective value of the current solution , determine the optimal solution , determine the objective value of the optimal solution , determine the number of iterations , determine the current temperature . Among them, is the preset initial temperature .

[0209] According to , construct the operation plan of the neighboring trains

[0210] Convert the eigenvectors of all trains in the operation plan of the neighboring trains into multi - channel data through the two - dimensional wavelet packet decomposition module . Among them, is the operation plan of the neighboring trains constructed in the - th iteration

[0211] Based on the convolutional neural network and the occupancy rate of each train, perform a pre - evaluation of the transportation efficiency to obtain the efficiency evaluation value . .

[0212] Determine the evaluation difference .

[0213] If , then update , update , update , update . If but , then update , update . Among them, is the natural exponential function, is the random function

[0214] .

[0215] If it is determined that the termination condition is not met according to and , then re - execute the steps of constructing the operation plan of the neighboring trains according to and the subsequent steps. If it is determined that the termination condition is met according to and , then determine that the optimized operation plan is , and the transportation efficiency evaluation value of the optimized operation plan is 。

[0216] Among them, according to Construct the operation plan of the neighborhood trains, including: For Perform the following processing on the target trains in To obtain the operation plan of the neighborhood trains: Suspend the operation of the first target train in . Among them, the first target train is a single - group train, and its passenger occupancy rate is lower than ,

[0217] Adjust the second target train in To a long - formation train with a smaller seating capacity with a probability of . Among them, Is the preset first probability value. The second target train is a long - formation train or a multiple - unit train, and its passenger occupancy rate is within , Is the preset second ratio value, Is the preset third ratio value, .

[0218] Adjust the third target train in To a single - group train with a probability of . Among them, Is the preset second probability value, . The third target train is a long - formation train or a multiple - unit train, and its passenger occupancy rate is lower than .

[0219] Add The trains corresponding to the fourth target train in . Among them, the fourth target train is a long - formation train or a multiple - unit train or a short - formation train that cannot perform formation increase, and its passenger occupancy rate is higher than ,

[0220] Adjust The arrival and departure times along the route of the fifth target train in . Among them, Is the train identification, Is the operation plan, Is the station The lower limit of the service frequency, Is the train And the station The associated identification, the train At the station If it stops at a station , otherwise .

[0221] Adjust according to the interval passing capacity constraint the operating section of the sixth target train. Among them, the passenger occupancy rate of the sixth target train is lower than , which is a preset fifth ratio value. The interval passing capacity constraint is: ; among them, is the interval identifier, , is the set of intervals of the railway network, is the number of trains entering the interval throughout the day, is the interval 's upper limit of passing capacity.

[0222] Among them, the objective function during transportation efficiency evaluation is: .

[0223] Among them, is the train identifier, is the operation plan, is 's target value, is the total number of stations passed by train , is the identifier of the stations passed by train , is the seating capacity of train , is the mileage of the interval , is the interval identifier, is the train For the interval of the th station, , is the station identifier, is the th station along the running route of train in the operation plan, is the th station along the running route of train in the operation plan, The passenger occupancy rate of train .

[0224] The constraint conditions during preliminary evaluation of transportation efficiency are: transport capacity allocation constraint, interval passing capacity constraint, number of trains starting and ending at stations constraint, basic diagram constraint, starting and ending time constraint, train seating capacity constraint, service frequency constraint.

[0225] Among them, the transport capacity allocation constraint is: . Among them, is the accounting coefficient of the kilometer number of the train , is the upper limit of the railway network transport capacity allocation.

[0226] The section passing capacity constraint is: . Among them, , is the set of sections of the railway network, is the number of trains entering section throughout the day, is the upper limit of the passing capacity of section .

[0227] The constraint on the number of trains originating and terminating at stations is: , . Among them, , is the set of stations in the railway network, is the first station along the running route of train in the operation plan, is the th station along the running route of train in the operation plan, is the upper limit of the originating and terminating capacity of station .

[0228] The basic diagram constraint is: . Among them, is the set of trains in the basic operation diagram, .

[0229] The originating and terminating time constraint is: . Among them, is the starting time of the all-day operation period, is the ending time of the all-day operation period, , is the originating time of train , is the arrival time of train at the th station.

[0230] The train seating capacity constraint is: . Among them, is the set of seating capacities of the equipped vehicle types, .

[0231] The service frequency constraint is: . Among them, is the lower limit of the service frequency of station , is the train Association identifier with the station For the train At the station If it stops Otherwise .

[0232] Among them, if , it is determined that the termination condition is not met. Among them, Is the preset total number of iterations.

[0233] If , then update , update . If , it is determined that the termination condition is not met. If , it is determined that the termination condition is met. Among them, Is the temperature drop ratio, Is the preset termination temperature.

[0234] Among them, the convolutional neural network includes: an input layer, a hidden layer, and an output layer.

[0235] The hidden layer includes: a convolutional layer, a pooling layer, and a fully connected layer.

[0236] There are 3 convolutional layers, and each convolutional layer has Convolution kernels, and the activation function is a rectified linear unit.

[0237] The pooling function of the pooling layer is an average pooling function.

[0238] The loss function of the convolutional neural network is a mean square error function, and the optimization function is the Adam algorithm.

[0239] Among them, the process of formulating the operation plan for high-speed railway trains is as follows: Determine the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0240] Based on the basic data of the high-speed railway network, determine the basic constraints of the high-feasibility high-speed railway operation plan verification HHTPV framework.

[0241] Initialize the HHTPV framework based on the constraint conditions. Among them, the constraint conditions are determined according to the basic constraints and the dynamic threshold constraint set.

[0242] Generate the operation plan in the current state according to the optimization strategy, the preferred adjustment direction of the operation plan, and the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0243] Within the HHTPV framework, verify the set of operation plans in the current state, and formulate the operation plan for high-speed railway trains according to the verification results.

[0244] On the basis of reasonable verification of the train operation plan, the generation of the railway passenger transport timetable can be carried out, which is equivalent to the optimization of the whole chain. The main considerations for the generation of the railway passenger transport timetable are as follows: The passenger flow information of the current station. Through the passenger flow information of the current station, the passenger flow heat is learned to obtain the passenger flow heat of the current station, the local area where the current station is located is determined, and the local heat matrix is constructed according to the passenger flow heat of other stations corresponding to the local area where the current station is located.

[0245] Perform eigenvalue decomposition on the local heat matrix to obtain the heat learning features corresponding to the local area where the current station is located, and generate the initial passenger transport timetable of the current station through the heat learning features. And determine the in-station dispatching coupling degree of the current station through the passenger flow heat and the passenger transport adjustment ratio of the current station.

[0246] Determining the in-station dispatching coupling degree of the current station specifically includes: Based on the preset coupling degree extraction period, trend values are respectively collected for the passenger flow heat change trend and the passenger transport adjustment trend to obtain a passenger flow heat trend value sequence and a passenger transport adjustment trend value sequence; After normalizing the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence, perform coupling degree analysis according to the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence to obtain the in-station dispatching coupling degree of the current station.

[0247] In specific implementation, the coupling degree extraction period can be obtained by mapping based on the current passenger flow density. In some embodiments, the coupling degree extraction period can be fixed as a constant value of 6h, and then the Pearson correlation coefficient between the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence can be used as the in-station dispatching coupling degree.

[0248] Dynamically feedback and update the initial passenger transport timetable according to the in-station dispatching coupling degree and the inter-station dispatching coupling degree to generate an optimized passenger transport timetable.

[0249] Aiming at the problems of insufficient dynamic response, difficult multi-professional collaboration and low calculation efficiency caused by the independent operation of the passenger flow prediction, train adjustment, seat allocation and timetable compilation links in the traditional method, a multi-professional modular collaboration system for passenger transport, transportation, dispatching and vehicles is constructed, a closed-loop process of "adjustment plan generation → timetable dynamic generation → operation evaluation → feedback optimization" is established, and lightweight data interaction is realized by using a standardized interface.

[0250] The device provided in this embodiment accurately evaluates the transportation efficiency according to the train operation plan of high-speed railway trains and the occupancy rate of each train, and then obtains an optimized operation plan, realizing the automatic optimization of the transportation efficiency evaluation of the high-speed railway train operation plan, and improving the optimization efficiency and accuracy.

[0251] Based on the same inventive concept of the optimization method for evaluating the transportation efficiency of high-speed railway train operation plans, this embodiment provides an electronic device, which is as Figure 5 shown, including: a memory 501, a processor 502, and a computer program.

[0252] Among them, the computer program is stored in the memory 501 and is configured to be executed by the processor 502 to implement the above-mentioned optimization method for evaluating the transportation efficiency of high-speed railway train operation plans.

[0253] Specifically, Obtain the train operation plan of high-speed railway trains and the occupancy rate of each train.

[0254] According to the train operation plan of high-speed railway trains and the occupancy rate of each train, determine the optimized operation plan and its transportation efficiency evaluation value through an evaluation model.

[0255] Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network.

[0256] The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data.

[0257] The convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rate of each train.

[0258] Among them, the convolutional neural network is trained based on historical multiple operation plans and their train occupancy rates to construct a mapping relationship for transportation efficiency evaluation.

[0259] Among them, according to the train operation plan of high-speed railway trains and the occupancy rate of each train, determining the optimized train operation plan of high-speed railway trains and its transportation efficiency evaluation value through the evaluation model includes: Convert the feature vectors of all trains in the train operation plan of high-speed railway trains into multi-channel data through the two-dimensional wavelet packet decomposition module . Among them, is the train operation plan of high-speed railway trains.

[0260] Through the convolutional neural network, based on and the occupancy rate of each train, conduct a preliminary evaluation of the transportation efficiency to obtain the efficiency evaluation value . Determine the current solution , and determine the objective value of the current solution , determine the optimal solution , determine the objective value of the optimal solution , determine the number of iterations , determine the current temperature . Among them, is the preset initial temperature.

[0261] According to construct the operation plan of the neighborhood trains.

[0262] Convert the feature vectors of all trains in the operation plan of the neighborhood trains into multi-channel data through the two-dimensional wavelet packet decomposition module . Among them, is the operation plan of the neighborhood trains constructed in the th iteration.

[0263] Through the convolutional neural network, based on and the occupancy rate of each train, pre-evaluate the transportation efficiency and obtain the efficiency evaluation value .

[0264] Determine the evaluation difference .

[0265] If , then update , update , update , update . If but , then update , update . Among them, is the natural exponential function, is the random function.

[0266] .

[0267] If it is determined according to and that the termination condition is not satisfied, then re-execute the steps of constructing the operation plan of the neighborhood trains according to and the subsequent steps. If it is determined according to and that the termination condition is satisfied, then determine that the optimized operation plan is , and the transportation efficiency evaluation value of the optimized operation plan is .

[0268] Among them, constructing the operation plan of the neighborhood trains according to includes:[[]] Perform the following processing on the target train in to obtain the operation plan of the neighborhood trains: Out of service the first target train in. Among them, the first target train is a single - group train, and its passenger occupancy rate is lower than , which is a preset first ratio value.

[0269] Adjust the second target train in to a long - formation train with a smaller seating capacity with a probability of . Among them, is a preset first probability value. The second target train is a long - formation train or a coupled train, and its passenger occupancy rate is within , is a preset second ratio value, is a preset third ratio value, .

[0270] Adjust the third target train in to a single - group train with a probability of . Among them, is a preset second probability value, . The third target train is a long - formation train or a coupled train, and its passenger occupancy rate is lower than .

[0271] Increase the train corresponding to the fourth target train in. Among them, the fourth target train is a long - formation train or a coupled train or a short - formation train that cannot perform formation increase, and its passenger occupancy rate is higher than , is a preset fourth ratio value, and at the same time, there is a peak - line train pair corresponding to the high - utilization section of the fourth target train at adjacent time points in the basic diagram.

[0272] Adjust the arrival and departure times of the fifth target train along the line in. Among them, the stations along the line of the fifth target train include the target station, and the target station satisfies the service frequency constraint. The service frequency constraint is: ; among them, is the train identification, is the operation plan, is the station 's service frequency lower limit, is the association identification between the train and the station . If the train stops at the station , then , otherwise .

[0273] Adjust according to the interval passing - capacity constraint The operating section of the sixth target train. Among them, the passenger occupancy rate of the sixth target train is lower than , which is a preset fifth ratio value. The interval passing capacity constraint is: ; among them, is the interval identifier, , is the set of intervals of the railway network, is the number of trains entering the interval throughout the day, is the upper limit of the passing capacity of the interval .

[0274] Among them, the objective function during the transportation efficiency evaluation is: .

[0275] Among them, is the train identifier, is the operation plan, is the target value, is the total number of stations passed by train , is the station identifier passed by train , is the capacity of train , is the mileage of the interval , is the interval identifier, is the For the interval of the th station, , is the station identifier, is the th station along the passing route of train in the operation plan, is the th station along the passing route of train in the operation plan, The passenger occupancy rate of the train.

[0276] The constraint conditions during the preliminary evaluation of transportation efficiency are: transportation capacity allocation constraint, interval passing capacity constraint, number of trains starting and ending at stations constraint, basic diagram constraint, starting and ending time constraint, train capacity constraint, service frequency constraint.

[0277] Among them, the transportation capacity allocation constraint is: . Among them, is the accounting coefficient of the kilometer number of train , is the upper limit of the railway network transportation capacity allocation.

[0278] The capacity constraint for sections is: . Among them, , is the set of sections in the railway network, is the number of trains entering section throughout the day, is the upper limit of the passing capacity of section .

[0279] The constraint for the number of trains originating and terminating at stations is: , . Among them, , is the set of stations in the railway network, is the first station along the running route of train in the operation plan, is the -th station along the running route of train in the operation plan, is the upper limit of the originating and terminating capacity of station .

[0280] The constraint for the basic train diagram is: . Among them, is the set of trains in the basic operation diagram, .

[0281] The constraint for the originating and terminating times is: . Among them, is the starting time of the whole-day operation period, is the ending time of the whole-day operation period, , is the originating time of train , is the arrival time of train at the -th station.

[0282] The constraint for the train capacity is: . Among them, is the set of capacities of the equipped train types, .

[0283] The constraint for the service frequency is: . Among them, is the lower limit of the service frequency of station , is the association identifier between train and station . If train stops at station , then , otherwise .

[0284] Among them, if , it is determined that the termination condition is not satisfied. Among them, is the preset total number of iterations.

[0285] If , then update , update . If , it is determined that the termination condition is not satisfied. If , it is determined that the termination condition is satisfied. Among them, is the temperature drop ratio, is the preset termination temperature.

[0286] Among them, the convolutional neural network includes: an input layer, a hidden layer, and an output layer.

[0287] The hidden layer includes: a convolutional layer, a pooling layer, and a fully connected layer.

[0288] There are 3 convolutional layers, and each convolutional layer has convolution kernels, and the activation function is the rectified linear unit.

[0289] The pooling function of the pooling layer is the average pooling function.

[0290] The loss function of the convolutional neural network is the mean squared error function, and the optimization function is the Adam algorithm.

[0291] Among them, the process of formulating the operation plan of high-speed railway trains is as follows: Determine the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0292] Based on the basic data of the high-speed railway network, determine the basic constraints of the high-feasibility high-speed rail operation plan verification HHTPV framework.

[0293] Initialize the HHTPV framework based on the constraint conditions. Among them, the constraint conditions are determined according to the basic constraints and the dynamic threshold constraint set.

[0294] Generate the operation plan in the current state according to the optimization strategy, the preferred adjustment direction of the operation plan, and the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0295] Within the HHTPV framework, verify the set of operation plans in the current state, and formulate the operation plan of high-speed railway trains according to the verification results.

[0296] On the basis of reasonable verification of the operation plan formulation, railway passenger train timetables can be generated, which is equivalent to full-chain optimization. The main factors considered in generating railway passenger train timetables: The passenger flow information of the current station is used to learn the passenger flow heat of the current station through the passenger flow information of the current station, obtain the passenger flow heat of the current station, determine the local area where the current station is located, and construct a local heat matrix based on the passenger flow heat of other stations corresponding to the local area where the current station is located.

[0297] Perform eigenvalue decomposition on the local heat matrix to obtain the heat learning features corresponding to the local area where the current station is located, and generate the initial passenger transport timetable of the current station through the heat learning features. And determine the in-station scheduling coupling degree of the current station through the passenger flow heat and the passenger transport adjustment ratio of the current station.

[0298] Determining the in-station scheduling coupling degree of the current station specifically includes: Based on a preset coupling degree extraction period, respectively collect trend values of the passenger flow heat change trend and the passenger transport adjustment trend to obtain a passenger flow heat trend value sequence and a passenger transport adjustment trend value sequence; After normalizing the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence, perform coupling degree analysis according to the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence to obtain the in-station scheduling coupling degree of the current station.

[0299] When specifically implemented, the coupling degree extraction period can be obtained by mapping based on the current passenger flow density. In some embodiments, the coupling degree extraction period can be fixed as a constant value of 6h, and then the Pearson correlation coefficient between the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence can be used as the in-station scheduling coupling degree.

[0300] Dynamically feedback and update the initial passenger transport timetable according to the in-station scheduling coupling degree and the inter-station scheduling coupling degree to generate an optimized passenger transport timetable.

[0301] Aiming at the problems of insufficient dynamic response, difficult multi-professional collaboration, and low calculation efficiency caused by the independent operation of passenger flow prediction, train adjustment, seat allocation, and timetable compilation in traditional methods, a multi-professional modular collaboration system for passenger transport, transportation, dispatching, and vehicles is constructed, a closed-loop process of "adjustment plan generation → timetable dynamic generation → operation evaluation → feedback optimization" is established, and lightweight data interaction is realized by using a standardized interface.

[0302] The electronic device provided in this embodiment, on which the computer program is executed by the processor to accurately evaluate the transport efficiency according to the train operation plan of high-speed railway trains and the passenger occupancy rate of each train, and then obtain an optimized operation plan, realizing the automatic optimization of the transport efficiency evaluation of the high-speed railway train operation plan, and improving the optimization efficiency and accuracy.

[0303] Based on the same inventive concept of the optimization method for evaluating the transportation efficiency of the high-speed rail train operation plan, this embodiment provides a computer-readable storage medium, on which a computer program is stored. The computer program is executed by a processor to implement the above-mentioned optimization method for evaluating the transportation efficiency of the high-speed rail train operation plan.

[0304] Specifically, Obtain the operation plan of high-speed rail trains and the occupancy rates of each train.

[0305] According to the operation plan of high-speed rail trains and the occupancy rates of each train, determine the optimized operation plan and its transportation efficiency evaluation value through an evaluation model.

[0306] Among them, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network.

[0307] The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the feature vectors of all trains in the operation plan into multi-channel data.

[0308] The convolutional neural network is used to evaluate the transportation efficiency based on the multi-channel data and the occupancy rates of each train.

[0309] Among them, the convolutional neural network is trained based on multiple historical operation plans and their train occupancy rates to construct a mapping relationship for transportation efficiency evaluation.

[0310] Among them, according to the operation plan of high-speed rail trains and the occupancy rates of each train, determining the optimized operation plan of high-speed rail trains and its transportation efficiency evaluation value through the evaluation model includes: Convert the feature vectors of all trains in the operation plan of high-speed rail trains into multi-channel data through the two-dimensional wavelet packet decomposition module . Among them, is the operation plan of high-speed rail trains.

[0311] Through the convolutional neural network, based on and the occupancy rates of each train, conduct a preliminary evaluation of the transportation efficiency to obtain an efficiency evaluation value . Determine the current solution , determine the objective value of the current solution , determine the optimal solution , determine the objective value of the optimal solution , determine the number of iterations , determine the current temperature . Among them, is the preset initial temperature.

[0312] According to Construct the operation plan of neighboring trains.

[0313] Convert the eigenvectors of all trains in the operation plan of neighboring trains into multi-channel data through a two-dimensional wavelet packet decomposition module . Among them, is the operation plan of neighboring trains constructed for the -th iteration.

[0314] Through a convolutional neural network, based on and the passenger occupancy rate of each train, pre-evaluate the transportation efficiency to obtain an efficiency evaluation value .

[0315] Determine the evaluation difference .

[0316] If , then update , update , update , update . If but , then update , update . Among them, is the natural exponential function, is the random function.

[0317] .

[0318] If it is determined that the termination condition is not satisfied according to and , then re-execute the steps of constructing the operation plan of neighboring trains according to and the subsequent steps. If it is determined that the termination condition is satisfied according to and , then determine that the optimized operation plan is , and the transportation efficiency evaluation value of the optimized operation plan is .

[0319] Among them, constructing the operation plan of neighboring trains according to includes:[[]] Perform the following processing on the target trains in to obtain the operation plan of neighboring trains:[[]] Suspend operation of the first target train in. Among them, the first target train is a single-group train, and its passenger occupancy rate is lower than , is the preset first proportional value.[[]]

[0320] Adjust the second target train in to a long formation train with a smaller seating capacity with a probability of . Among them, is a preset first probability value. The second target train is a long train or a coupled train, and its passenger occupancy rate is within and is a preset second ratio value, is a preset third ratio value, .

[0321] Adjust the third target train in to a single - unit train with a probability of . Among them, is a preset second probability value, . The third target train is a long train or a coupled train, and its passenger occupancy rate is lower than .

[0322] Add the train corresponding to the fourth target train in. Among them, the fourth target train is a long train or a coupled train or a short - unit train that cannot perform formation increase, and its passenger occupancy rate is higher than , is a preset fourth ratio value. At the same time, there is a peak - line train pair corresponding to the high - utilization section of the fourth target train at adjacent time points in the basic diagram.

[0323] Adjust the arrival and departure times of the fifth target train along the line. Among them, the stations along the line of the fifth target train include the target station, and the target station satisfies the service - frequency constraint. The service - frequency constraint is: ; among them, is the train identification, is the operation plan, is the station 's service - frequency lower limit, is the train 's association identification with the station . If the train stops at the station , then , otherwise .

[0324] Adjust the operation section of the sixth target train in according to the interval - passing - capacity constraint. Among them, the passenger occupancy rate of the sixth target train is lower than , is a preset fifth ratio value. The interval - passing - capacity constraint is: ; among them, is the interval identification, , is the set of intervals of the railway network, is the number of trains entering the interval throughout the day, is the interval The upper limit of passing capacity.

[0325] Among them, the objective function during transportation efficiency evaluation is: .

[0326] Among them, is the train identification, is the operation plan, is the target value of, is the train total number of stations passed by, is the train identification of stations passed by, is the train seating capacity, is the mileage of the section , is the section identification, is the train For the section of the station, , is the station identification, is the th station along the running route of the train in the operation plan, , is the th station along the running route of the train in the operation plan, , The occupancy rate of the train.

[0327] The constraint conditions during pre-evaluation of transportation efficiency are: transportation capacity allocation constraint, section passing capacity constraint, number of trains starting and ending at stations constraint, basic diagram constraint, starting and ending time constraint, train seating capacity constraint, service frequency constraint.

[0328] Among them, the transportation capacity allocation constraint is: . Among them, is the accounting coefficient of the kilometer number of the train, is the upper limit of railway network transportation capacity allocation.

[0329] The section passing capacity constraint is: . Among them, , is the set of sections of the railway network, is the number of trains entering the section throughout the day, is the upper limit of passing capacity of the section.

[0330] The constraint on the number of trains starting and ending at stations is: , . Among them, , is the set of stations in the railway network, is the first station along the running route of the train in the train operation plan . is the first station along the running route of the train in the train operation plan is the th station is the upper limit of the origin-destination capacity of station .

[0331] The basic diagram constraint is: . Among them, is the set of trains in the basic operation diagram, .

[0332] The origin-destination time constraint is: . Among them, is the start time of the whole-day operation period, is the end time of the whole-day operation period, , is the departure time of train , is the arrival time of train at the th station

[0333] The train seating capacity constraint is: . Among them, is the set of seating capacities of the equipped train types, .

[0334] The service frequency constraint is: . Among them, is the lower limit of the service frequency of station , is the association identifier between train and station . If train stops at station , then , otherwise .

[0335] Among them, if , it is determined that the termination condition is not met. Among them, is the preset total number of iterations.

[0336] If , then update , update . If , it is determined that the termination condition is not met. If , it is determined that the termination condition is met. Among them, is the temperature drop ratio, is the preset termination temperature.

[0337] Among them, the convolutional neural network includes: an input layer, a hidden layer, and an output layer.

[0338] The hidden layer includes: a convolutional layer, a pooling layer, and a fully connected layer.

[0339] There are 3 convolutional layers, and each convolutional layer has convolution kernels, and the activation function is a rectified linear unit.

[0340] The pooling function of the pooling layer is an average pooling function.

[0341] The loss function of the convolutional neural network is a mean square error function, and the optimization function is the Adam algorithm.

[0342] Among them, the compilation process of the operation plan of high-speed railway trains is as follows: Determine the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0343] Based on the basic data of the high-speed railway network, determine the basic constraints of the high-feasibility high-speed rail operation plan verification HHTPV framework.

[0344] Initialize the HHTPV framework based on the constraint conditions. Among them, the constraint conditions are determined according to the basic constraints and the dynamic threshold constraint set.

[0345] Generate the operation plan in the current state according to the optimization strategy, the preferred adjustment direction of the operation plan, and the complete set of operation plans for passenger trains on the specified lines or regions in the next diagram period.

[0346] Within the HHTPV framework, verify the set of operation plans in the current state, and compile the operation plan of high-speed railway trains according to the verification results.

[0347] On the basis of reasonable verification of the compilation of the operation plan, railway passenger train timetables can be generated, which is equivalent to full-chain optimization. The main factors considered in the generation of railway passenger train timetables are: The passenger flow information of the current station. Through the passenger flow information of the current station, passenger flow heat learning is carried out to obtain the passenger flow heat of the current station, determine the local area where the current station is located, and construct a local heat matrix according to the passenger flow heat of other stations corresponding to the local area where the current station is located.

[0348] Perform eigenvalue decomposition on the local heat matrix to obtain the heat learning features corresponding to the local area where the current station is located, and generate the initial passenger train timetable of the current station through the heat learning features. And determine the in-station dispatching coupling degree of the current station according to the passenger flow heat and the passenger transport adjustment ratio of the current station.

[0349] Determining the in-station dispatching coupling degree of the current station specifically includes: Based on a preset coupling degree extraction period, trend value collection is respectively performed on the passenger flow heat change trend and the passenger transport adjustment trend to obtain a passenger flow heat trend value sequence and a passenger transport adjustment trend value sequence; After normalizing the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence, coupling degree analysis is performed according to the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence to obtain the in-station dispatching coupling degree of the current station.

[0350] When specifically implemented, the coupling degree extraction period can be mapped based on the current passenger flow density. In some embodiments, the coupling degree extraction period can be fixed as a constant value of 6h, and then the Pearson correlation coefficient between the passenger flow heat trend value sequence and the passenger transport adjustment trend value sequence can be used as the in-station dispatching coupling degree.

[0351] According to the in-station dispatching coupling degree and the inter-station dispatching coupling degree, dynamic feedback update is performed on the initial passenger transport timetable to generate an optimized passenger transport timetable.

[0352] Aiming at problems such as insufficient dynamic response, difficult multi-professional collaboration, and low calculation efficiency caused by the independent operation of passenger flow prediction, train adjustment, seat allocation, and timetable compilation links in traditional methods, a multi-professional modular collaboration system for passenger transport, transportation, dispatching, and vehicles is constructed, a closed-loop process of "adjustment plan generation → timetable dynamic generation → operation evaluation → feedback optimization" is established, and lightweight data interaction is realized by using standardized interfaces.

[0353] The computer-readable storage medium provided in this embodiment, on which the computer program is executed by a processor to accurately evaluate the transport efficiency according to the train operation plan of high-speed railway trains and the passenger occupancy rate of each train, and then obtain an optimized operation plan, realizes the automatic optimization of the transport efficiency evaluation of the high-speed railway train operation plan, and improves the optimization efficiency and accuracy.

[0354] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0355] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or one block or multiple blocks.

[0356] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or one block or multiple blocks.

[0357] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or one block or multiple blocks.

[0358] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, "a plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0359] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0360] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalent technologies, this application is also intended to include these modifications and variations.

Claims

1. A method for optimizing the transport efficiency evaluation of a high-speed train operation plan, characterized in that: The method comprises: Obtain the operation plan of high-speed railway trains and the passenger occupancy rate of each train; According to the operation plan of high-speed railway trains and the passenger occupancy rate of each train, the optimized operation plan and its transportation efficiency evaluation value are determined through the evaluation model; Wherein, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the characteristic vectors of all trains in the operation plan into multi-channel data; The convolutional neural network is used to evaluate the transportation efficiency based on multi-channel data and the passenger load factor of each train; The convolutional neural network is trained based on multiple historical operation plans and train occupancy rates to construct a mapping relationship and conduct transportation efficiency evaluation.

2. The method according to claim 1, characterized in that The step of determining the optimized high-speed railway train operation plan and its transport efficiency evaluation value through an evaluation model according to the high-speed railway train operation plan and the passenger occupancy rate of each train includes: The feature vectors of all trains in the high-speed railway train operation plan are converted into multi-channel data through a two-dimensional wavelet packet decomposition module. ;in, A plan for operating the high-speed railway trains; Through convolutional neural network, based on Carry out a preliminary evaluation of the transport efficiency based on the passenger load factor of each train and obtain the efficiency evaluation value ; Determine the current solution , determine the target value of the current solution , determine the optimal solution , determine the target value of the optimal solution , determine the number of iterations , determine the current temperature ;in, is the preset initial temperature; according to Construct the operation plan of neighborhood trains; The feature vectors of all trains in the neighborhood train operation plan are converted into multi-channel data through a two-dimensional wavelet packet decomposition module. ;in, For the The first iteration is the operation plan of the constructed neighborhood trains; Through convolutional neural network, based on Carry out a preliminary evaluation of the transport efficiency based on the passenger load factor of each train and obtain the efficiency evaluation value ; Determine the evaluation difference ; like , then update ,renew ,renew ,renew ;like but , then update ,renew ;in, is the natural exponential function, is a random function; ； If based on and If the termination condition is not met, the Steps and subsequent steps for constructing the operation plan of the neighborhood train; if according to and If the termination condition is met, the optimized operation plan is , and the transport efficiency evaluation value of the optimized operation plan is .

3. The method according to claim 2, characterized in that The basis Construct a neighborhood train operation plan, including: right The target train in is processed as follows to obtain the operation plan of the neighboring trains: Outage The first target train is a single train, and its passenger load factor is lower than , is a preset first ratio value; Will The second target train is The probability of adjusting to a longer train with a smaller capacity; among them, is the preset first probability value; the second target train is a long train or a multiple train, and its passenger occupancy rate is Inside, is the preset second ratio value, is the preset third ratio value, ; Will The third target train is The probability of is adjusted to a single train; among them, is a preset second probability value, The third target train is a long train or a multiple train, and its passenger load factor is lower than ; Increase The train corresponding to the fourth target train in the above formula; the fourth target train is a long train or a double-unit train or a short train that cannot increase the marshaling, and its passenger load factor is higher than , is a preset fourth ratio value, and at the same time, there is a peak line pair corresponding to the fourth target train high utilization section at adjacent time points in the basic diagram; Adjustment The arrival and departure time of the fifth target train along the way; wherein the stations along the way of the fifth target train include the target station, and the target station satisfies the service frequency constraint; the service frequency constraint is: ;in, For train identification, To launch the plan, For the station The lower limit of service frequency, For trains With the station The associated identifier of the train At the station Stop ,otherwise ; Adjust according to the interval passing capacity constraint The section where the sixth target train runs; among them, the passenger occupancy rate of the sixth target train is lower than , is the preset fifth ratio value; the interval passing capacity constraint is: ;in, is the interval identifier, , is the interval set of the railway network, For full-day access The number of trains, For interval The passing capacity limit.

4. The method according to claim 1, characterized in that The objective function for transport efficiency evaluation is: ; in, For train identification, To launch the plan, for The target value of For trains The total number of stations along the route, For trains The station sign of the route, For trains Capacity, For interval Mileage, is the interval identifier, For trains For The station area, , For station identification, For the trains in the operation plan Open the first Stations, For the trains in the operation plan Open the first Stations, train The passenger load factor; The constraints for transport efficiency pre-assessment are: capacity allocation constraints, interval throughput constraints, station departure and arrival train number constraints, basic diagram constraints, departure and arrival time constraints, train capacity constraints, and service frequency constraints. Among them, the capacity allocation constraint is: ;in, For trains The calculation coefficient of kilometers, capping the rail network’s capacity; The interval passing capacity constraint is: ;in, , is the interval set of the railway network, For full-day access The number of trains, For interval The upper limit of the passing capacity; The number of trains that originate and arrive at a station is constrained as follows: , ;in, , is the set of stations in the railway network, For the trains in the operation plan The first station along the route, For the trains in the operation plan Open the first Stations, For the station The starting and ending capacity limit; The basic graph constraints are: ;;in, is the train set of the basic operation diagram, ; The start and end time constraints are: ;in, The starting time of the full-day operation period. The end time of the full-day operation period. , For trains The departure time, For trains In the The arrival time of each station; The train capacity constraint is: ;in, To equip the vehicle model with a fixed number of people, ; The service frequency constraint is: ;in, For the station The lower limit of service frequency, For trains With the station The associated identifier of the train At the station Stop ,otherwise .

5. The method according to claim 2, characterized in that: like , it is determined that the termination condition is not met; among them, is the preset total number of iterations; like , then update ,renew ;like , then it is determined that the termination condition is not met. If , then it is determined that the termination condition is met; among them, is the temperature drop ratio, is the preset end temperature.

6. The method according to claim 1 or 2, characterized in that: The convolutional neural network comprises: an input layer, a hidden layer and an output layer; The hidden layer includes: a convolutional layer, a pooling layer and a fully connected layer; There are three convolutional layers, each of which has The convolution kernel of is, and the activation function is a linear rectifier unit; The pooling function of the pooling layer is an average pooling function; The loss function of the convolutional neural network is the mean square error function, and the optimization function is the Adam algorithm.

7. The method according to claim 1, characterized in that The process of compiling the operation plan of high-speed railway trains is as follows: Determine the complete set of passenger train operation plans for the designated lines or areas in the next chart period; Based on the basic data of the high-speed railway network, a highly feasible high-speed railway operation plan is determined to verify the basic constraints of the HHTPV framework; Initializing the HHTPV framework based on constraint conditions; wherein the constraint conditions are determined according to the basic constraint and the dynamic threshold constraint set; Generate a running plan in the current state according to the optimization strategy, the optimization adjustment direction of the running plan and the complete set of passenger train running plans for the designated lines or areas in the next diagram period; Within the HHTPV framework, the set of operation plans under the current status is verified, and the operation plan of high-speed railway trains is compiled based on the verification results.

8. A high-speed train operation plan transportation efficiency evaluation and optimization device, characterized in that: The device comprises: An acquisition module is used to obtain the operation plan of high-speed railway trains and the passenger occupancy rate of each train; An optimization module is used to determine the optimized operation plan and its transport efficiency evaluation value through an evaluation model according to the operation plan of high-speed railway trains and the passenger occupancy rate of each train; Wherein, the evaluation model includes a two-dimensional wavelet packet decomposition module and a convolutional neural network; The decomposition function of the two-dimensional wavelet packet decomposition module is Daubechies4, and the two-dimensional wavelet packet decomposition module is used to convert the characteristic vectors of all trains in the operation plan into multi-channel data; The convolutional neural network is used to evaluate the transportation efficiency based on multi-channel data and the passenger load factor of each train; The convolutional neural network is trained based on multiple historical operation plans and train occupancy rates to construct a mapping relationship and conduct transportation efficiency evaluation.

9. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon; the computer program is executed by a processor to implement the method according to any one of claims 1 to 7.

Images