Virtual marshalling-oriented train working diagram self-adaptive adjustment method and virtual marshalling-oriented train working diagram self-adaptive adjustment system

Abstract

The invention discloses a virtual marshalling-oriented train working diagram adaptive adjustment method, which comprises the following steps of: monitoring and collecting original state data of train operation in real time, and calculating operation state parameters of virtual marshalling based on the original state data; forming a virtual marshalling operation characteristic constraint based on the operation state parameters of the virtual marshalling, and predicting an operation trend based on the virtual marshalling operation characteristic constraint to form an operation trend prediction result; dynamically reconstructing an operation permission boundary based on the marshalling state; performing multi-target adaptive optimization adjustment on the virtual marshalling train working diagram by taking the operation permission boundary parameter set as a constraint condition and combining the operation trend prediction result through a multi-target optimized working diagram adaptive adjustment algorithm, and generating an adjustment scheme as an optimized working diagram adjustment result; and generating an operation instruction based on the optimized operation diagram adjustment result, and performing cooperative control with the CTC / RBC based on the operation instruction. The invention further discloses a system, electronic equipment and a computer readable storage medium.

Description

Technical Field

[0001] This invention relates to the field of intelligent railway transportation organization and scheduling technology, specifically to a method and system for adaptively and dynamically adjusting train timetables for virtual train formation operation modes without changing the basic safety control logic of the train control system, and particularly to a method and system for adaptive adjustment of train timetables for virtual train formations. Background Technology

[0002] With the continuous expansion of railway transportation scale and the increasing complexity of transportation organization, train operation organization is gradually evolving from the traditional independent operation of single trains to multi-train collaborative operation. Virtual formation, as a new train operation mode based on train-to-train communication and collaborative control, has significant advantages in improving line capacity and operational efficiency by reducing train tracking intervals and supporting dynamic train formation and de-formation.

[0003] In virtual train formation operation mode, the longitudinal running relationship between multiple trains no longer depends on fixed block sections or static tracking intervals, but is closely related to factors such as the relative running status between trains, differences in braking performance, and communication link status. Its operational safety constraints exhibit significant dynamism and coupling. This operational characteristic necessitates that train operation permit boundaries and operational constraints change in real time with the operating status.

[0004] Existing centralized railway dispatching systems typically rely on pre-compiled train timetables, making local adjustments during operation via manual or semi-automatic methods. These adjustments are primarily based on fixed minimum tracking intervals, minimum time intervals between sections, and station capacity constraints. While this dispatching method is well-suited for traditional operating modes, it struggles to fully reflect the impact of changes in train formation structure and relative operating conditions on operational safety and efficiency under virtual train formation operation conditions. This can easily lead to conservative operational constraint settings or delayed dispatching adjustments.

[0005] On the other hand, the train control system can dynamically generate train operation permits based on train braking capacity and safety protection principles, but its focus is on the safety protection of a single train or adjacent trains, lacking a comprehensive consideration of the overall timetable adjustment objectives, and making it difficult to achieve coordinated optimization between operational efficiency and safety constraints from the perspective of transportation organization.

[0006] Therefore, without changing the existing safety control logic of the train control system, how to make full use of the virtual train formation operation status information, establish a coordination mechanism between the scheduling level operation diagram adjustment decision and the train control level operation permission boundary, and realize the adaptive dynamic adjustment of the train operation diagram has become an urgent technical problem to be solved under the virtual train formation operation condition. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method and system for adaptive adjustment of train timetables for virtual train formations. Without altering the existing safety control principles of the train control system, this invention dynamically adjusts the train timetable by introducing a virtual train formation operation status perception and operation trend prediction mechanism, thereby achieving coordination and consistency between scheduling decisions and operational safety constraints. Specifically, this invention acquires the operation and communication status information of trains within the virtual train formation, constructs a virtual train formation operation status description model, and predicts train operation trends based on this status information. On this basis, it dynamically reconstructs operation permission boundary parameters that match the virtual train formation operation status and introduces them as constraints into the train timetable adjustment process. Through a multi-objective optimization approach, it comprehensively considers train punctuality, operational smoothness, train formation stability, and operational conflict risks to achieve adaptive adjustment of the train timetable.

[0008] It should be noted that the method described in this invention primarily operates at the centralized dispatching system level, used to generate train timetable adjustment schemes and corresponding operational constraint parameters. The operational permission boundary parameters are suggested constraint conditions calculated by the dispatching layer, and their final effectiveness still requires verification and execution by the train control system based on existing safety control logic. Therefore, this invention does not involve any modification to the principles of train control system movement authorization generation, speed monitoring, or braking control.

[0009] The first aspect of the present invention is to provide an adaptive adjustment method for train timetables for virtual train formations, comprising the following steps:

[0010] S1, monitor and collect raw status data of train operation in real time, and calculate the operating status parameters of virtual train formation based on the raw status data;

[0011] S2, based on the operating status parameters of the virtual group, virtual group operating characteristic constraints are formed, and based on the virtual group operating characteristic constraints, operating trends are predicted to form operating trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operating permission boundaries and adjustment of the operating diagram;

[0012] S3 dynamically reconstructs the operational permission boundary based on the grouping state, thereby predicting and reconstructing operational constraints in real time and outputting the operational permission boundary parameter set, making the operation diagram adjustment "adaptive".

[0013] S4, using a multi-objective adaptive adjustment algorithm for the train schedule, with the set of operational permission boundary parameters output in S3 as constraints, and combined with the operational trend prediction results in S2, performs multi-objective adaptive optimization adjustment on the virtual train schedule, generating an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation as the optimized train schedule adjustment result; wherein, the multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods, and those skilled in the art can implement the corresponding solution process based on the above objective function and constraints;

[0014] S5, generate running instructions based on the optimized running diagram adjustment results, and perform coordinated control with CTC / RBC based on the running instructions.

[0015] Preferably, S1 includes:

[0016] S11, Collect raw status data of train operation, wherein the raw status data includes single-vehicle operation parameters, group train status and communication quality data;

[0017] Raw train operation status data is collected in real time through interfaces with the RBC and onboard systems, including:

[0018] (1) Single-vehicle operating parameters, including: train exist The position of the front of the car at that moment Rear of the car Longitudinal running speed Longitudinal acceleration and minimum emergency braking acceleration ;

[0019] (2) Group train status, including: relative distance between trains and relative velocity ;

[0020] (3) Communication quality data, including: vehicle-to-vehicle or vehicle-to-ground communication links. Communication strength at any given moment Packet loss rate Communication delay ;

[0021] S12, Calculate the relative distance and speed difference between trains based on the original state data;

[0022] For the car behind For example, it is related to the car in front. relative distance With speed difference As shown in equations (1) and (2) respectively:

[0023] (1);

[0024] (2);

[0025] S13, Evaluating link stability based on normalization function The normalization function is shown in equations (3) and (4) below:

[0026] (3);

[0027] (4);

[0028] in, This represents the communication signal strength; a higher value indicates a more stable link. and These represent the upper and lower limits of the signal strength, respectively. Normalized communication signal strength; , and The three weighting coefficients are three non-negative fixed constants that do not change with time. change; for The probability of link error, packet loss, or failure at any given moment; for The communication link delay, transmission delay, or response delay at any given moment;

[0029] S14, Calculate the dynamic safety distance, including:

[0030] (1) Calculate the emergency braking distance As shown in equation (5):

[0031] (5);

[0032] in, For minimum emergency braking acceleration, This refers to the longitudinal running speed;

[0033] (2) Calculate the impact compensation distance for velocity difference As shown in equation (6):

[0034] (6);

[0035] in, The impact compensation coefficient is used to reflect the longitudinal dynamic uncertainty in virtual train formation operation. Its value can be set according to the train type, train formation operation level or historical operation data.

[0036] (3) Calculate the communication and control delay compensation distance The calculation formula is shown in equation (7):

[0037] (7);

[0038] Among them, the overall system delay time Due to communication delay Controlling computation delay and execution response latency Together they constitute the whole, and their calculation formula is shown in equation (8):

[0039] (8);

[0040] in, The data transmission delay used to characterize vehicle-to-vehicle and vehicle-to-ground communication is calculated using the formula shown in equation (9):

[0041] (9);

[0042] in, and They represent the first The sending and receiving times of the next communication data packet To count the number of data packets within the window;

[0043] The calculation formula is used to characterize the calculation and processing time of the scheduling system or train control system for the running instructions, as shown in equation (10):

[0044] (10);

[0045] in, This indicates the time when the scheduling system or train control system receives the status data. Indicates the time when the corresponding operation control command was generated and completed;

[0046] The formula used to characterize the train-end response time is shown in equation (11):

[0047] (11);

[0048] in, This indicates the time when the train receives the control command. Indicates the time at which the train begins to perform traction, braking, or train formation control actions;

[0049] S15, Calculate dynamic safety distance As shown in equation (12):

[0050] (12);

[0051] The dynamic safety distance Unlike traditional safety distances based on fixed braking models;

[0052] S16, based on the dynamic safety distance Calculate the safety margin of virtual groups The calculation is shown in equation (13):

[0053] (13);

[0054] like This indicates a security instability and requires immediate triggering of S3 to tighten the permission boundaries;

[0055] S17, based on the virtual grouping security margin Constructing the state vector of the virtual group At any moment The state vector of the virtual group As shown in equation (14):

[0056] (14).

[0057] Preferably, S2 includes:

[0058] S21, Construct the predicted input features, including: at time... Based on the state vector of the virtual group Forming an input state feature sequence for trend prediction. As shown in equation (15):

[0059] (15);

[0060] in, The state sampling time interval, The length of the time series;

[0061] S22, a communication uncertainty constraint mechanism is introduced during the trend prediction process, thereby forming a weighted state sequence under communication uncertainty constraints, including: based on the stability of the communication link. For the input state feature sequence Different weights are assigned to historical states at different times to reduce the impact of communication jitter and latency fluctuations on the prediction results of operational trends;

[0062] S23, Based on the acceleration change characteristic quantity, the nonlinear operating characteristics caused by the ungrouping or regrouping of the virtual group are modeled, and the acceleration change characteristic quantity is introduced in the operation trend prediction process, as shown in equation (16):

[0063] (16);

[0064] in, The time step for running the trend prediction model, For longitudinal acceleration;

[0065] S24, Construct a running trend prediction model, including: based on the input state feature sequence A time-series prediction model with a gating structure is constructed as an operation trend prediction model. The operation trend prediction model is used to model the time-series evolution relationship of the virtual train formation operation state. The gating structure is used to comprehensively consider the relative operation state changes between trains in the virtual train formation, the impact of communication link stability on the effectiveness of historical information, and the disturbance effect of nonlinear acceleration changes on the operation trend during the de-forming and re-forming process during the state update process.

[0066] S25, Output the virtual grouping operation trend prediction result based on the operation trend prediction model. As shown in equation (17):

[0067] (17);

[0068] in, For train At the station The predicted arrival time; For train At the station Predicted departure times; For train In the interval Predicted runtime; Predict the probability of conflict during virtual grouping operations; A predictive score for the stability of virtual group operation.

[0069] Preferably, S3 includes:

[0070] S31, the parameter set of the operation permission boundary that constitutes the operation permission boundary parameter system is defined. Under the virtual grouping operation condition, the parameter set of the operation permission boundary consists of the following key parameters, as shown in equation (18):

[0071] (18);

[0072] in, Indicates the dynamic tracking interval; Indicates the minimum time minute for interval operation; Indicates the permitted operating distance for the train; This represents the set of executable overrun time windows; This indicates the set of sections or stations that can be grouped / degrouped; the above parameters together constitute the train's schedule. Permissible operating boundaries;

[0073] S32, regarding the dynamic tracking interval Perform multi-factor reconstruction, including: the virtual grouping safety margin obtained based on S15. The virtual grouping operation conflict prediction probability obtained by S2 and virtual grouping stability score The minimum safe tracking interval is dynamically reconstructed, as shown in equation (19):

[0074] (19);

[0075] in: The baseline tracking interval is set under ideal operating conditions. For virtual grouping safety margin; Predict the probability of conflict during virtual grouping operations; Assess the stability of virtual groupings; , , The weight coefficients are non-negative and ;

[0076] S33, Based on the physical length of the interval and the predicted interval running time, calculate the minimum time division of the interval. Dynamic correction is performed, as shown in equation (20):

[0077] (20);

[0078] in, For the time interval of the graph, The risk correction coefficient for the minimum time division of the interval is used to reflect the impact of communication uncertainty, changes in group stability, and operational conflict risk on the interval's operational safety margin during virtual grouping operation, as shown in equation (21):

[0079] (twenty one);

[0080] in, , and The weight coefficients are non-negative and ;

[0081] S34, the available operating distance of the train The recommended parameters for the operational clearance boundary are synchronized to the train control system, which then determines the available operational clearance distance for the train based on the dynamic safety distance and existing safety control logic. Conduct security checks and implement them at all times; The available operating distance of the train calculated by the scheduling layer The following constraint relationship is satisfied:

[0082] (twenty two);

[0083] in, The dynamic safety distance calculated in step S14; The maximum authorized position is limited by track conditions or the capabilities of the train control system;

[0084] S35, Filter Override and Group Resource Availability Window, including:

[0085] (1) Filter the set of overrun time windows As shown in equation (23):

[0086] (twenty three);

[0087] in: Indicates at the station At this point, the predicted time margin for train overtaking or decoupling operations under operating conditions is used to assess whether station operations have a feasible time window. This indicates a time conflict in station operations, making overtaking or decoupling impossible. The calculation formula is shown in equation (24):

[0088] (twenty four);

[0089] in, Trains predicted for S2 Arrival at the station The moment; Trains predicted for S2 Leaving the station The moment; To address prediction errors, communication jitter, and job execution uncertainties, a time safety redundancy is introduced, calculated as shown in equation (25):

[0090] (25);

[0091] in, The average time for station route arrangement and scheduling is calculated. The average transmission delay between scheduling instructions and train control commands can be reused from S14. ; The physical execution time for on-site route planning, turnout switching to signal opening; A fixed safety redundancy time configured for the system;

[0092] This is a track capacity indicator used to indicate whether station s has the capacity to handle the current virtual marshalling operation within the prediction time window. The determination condition is shown in equation (26):

[0093] (26);

[0094] The track length, track occupancy status, and route availability information of the station are provided by the CTC station basic database and real-time route management module;

[0095] (2) Establish a set of grouping / degrouping sections As shown in equation (27):

[0096] (27);

[0097] in: The communication link stability calculated for S13; This is the threshold for link stability; This refers to the length of available tracks at the station. This refers to the total length of the virtual train formation.

[0098] S36, Output runtime license boundaries, including: the runtime license boundary parameter set obtained from dynamic reconstruction. The output is sent to S4 as a dynamic constraint condition for the multi-objective adaptive adjustment model of the running graph, ensuring that the generated running graph adjustment scheme meets the requirements of virtual grouping operation in terms of security, executability and communication reliability.

[0099] Preferably, S4 includes:

[0100] S41, optimize the input and define the decision variables; including:

[0101] (1) Optimize input, including:

[0102] Optimize the dynamic runtime permission boundary, as shown in equation (28):

[0103] (28);

[0104] The optimized trend prediction output is shown in equation (29):

[0105] (29);

[0106] (2) Define decision variables, including:

[0107] Define the set of decision variables for running graph adjustment As shown in equation (30):

[0108] (30);

[0109] in, Indicates train At the station Adjustment amount for arrival and departure times; Indicates train At the station Stock market selection decisions; This represents the virtual group switching decision variable, where 1 indicates performing decompilation / recompilation;

[0110] S42, construct the comprehensive multi-objective optimization objective function, as shown in equation (31):

[0111] (31);

[0112] in, , , , These are the weighting coefficients, and =1;

[0113] For punctual targets, the calculation formula is shown in equation (32):

[0114] (32);

[0115] in, , , and The diagram shows the arrival and departure times, where... and These represent the adjusted trains. At the station Arrival time;

[0116] To achieve the smoothness target or energy consumption proxy target, the operating time adjustment range is used as a proxy indicator for energy consumption and comfort. The calculation formula is shown in equation (33):

[0117] (33);

[0118] in, and They represent trains The start and end times within the current optimization window; Indicates train At any moment The longitudinal acceleration of the movement; This represents the rate of change of the train's longitudinal acceleration over time. and This is the smoothness weighting coefficient; Used to constrain operating efficiency and energy consumption. Used to constrain operational comfort and longitudinal stability of train formation; the operational smoothness target or energy consumption proxy target is only used as an evaluation index for timetable adjustment and does not directly participate in train traction or braking control;

[0119] For the group stability target, the calculation is shown in equation (34):

[0120] (34);

[0121] To implement a conflict risk penalty, which guides the system to prioritize the operating scheme with lower conflict risk when multiple solutions are feasible, the calculation is shown in equation (35):

[0122] (35);

[0123] S43, define the constraints, including minimum time division constraint, dynamic tracking interval constraint, movement authorization constraint, and station resources and marshalling operation constraint;

[0124] The minimum time-division constraint of the interval is shown in equation (36):

[0125] (36);

[0126] When the optimizer finds that it cannot... When a feasible solution is found within the constraints, the system will trigger a priority inversion mechanism, prioritizing the sacrifice of the punctuality objective. To ensure a safety margin ;

[0127] The dynamic tracking interval constraint is shown in equation (37):

[0128] (37);

[0129] The mobility authorization constraints are as shown in equation (38):

[0130] (38);

[0131] The station resources and marshalling operation constraints are shown in equations (39) and (40):

[0132] (39);

[0133] (40);

[0134] S44, based on heuristic search and a rolling time-domain multi-objective optimization strategy, performs adaptive solution and rolling update, including:

[0135] (1) Initialization: based on the current state Compared with the prediction results Initial value;

[0136] (2) Window scrolling: Set up scrolling optimization for windows, in the window [ Internally solve for the optimal adjustment sequence;

[0137] (3) Conflict detection and resolution: If a conflict is detected Automatically increase the weight of safety and risk objectives;

[0138] (4) Scheme output: Output the executable operation diagram adjustment scheme at the current time;

[0139] (5) When a feasible solution that satisfies all constraints cannot be obtained within the current rolling optimization window, the system automatically reverts to the most recently executable running graph state and increases the weight of safety and risk-related objectives to solve the problem again.

[0140] S45, Output the optimized runtime chart adjustment results. And pass it to S5 for execution. The optimized runtime graph adjustment result is shown in equation (41):

[0141] (41);

[0142] in, This indicates the train's arrival time after adjustments at each station. This indicates the train's departure time after adjustments at each station. This indicates the decision-making process for train track occupancy at the station, the corresponding route usage window, and the range of the occupied section; This indicates a command to switch virtual group status.

[0143] Preferably, S5 includes:

[0144] S51, parse the run graph adjustment results and form an instruction parameter mapping, including: the optimized run graph adjustment results obtained in S4. The results of the runtime chart adjustment are analyzed in a structured manner, and the following three types of instruction parameters are generated as the input basis for the generation of runtime instructions:

[0145] (1) Time-related parameters, including: the adjusted arrival times of trains at each station. Adjusted departure times of trains at various stations ;

[0146] (2) Resource parameters, including: track occupancy decisions for trains at stations. And the corresponding route usage window and occupied section range;

[0147] (3) Grouping control parameters, including: virtual grouping status switching instructions This is used to instruct trains to form, detach, or maintain their current formation status.

[0148] S52, the centralized scheduling system generates scheduling layer operation instructions for the operation organization layer based on the parsing of the instruction parameters. Multiple scheduling layer operation instructions form a scheduling instruction set, including:

[0149] (1) Train timetable adjustment plan: used to update the arrival and departure times, overtaking relationships and track utilization plans in the train timetable, so that the scheduling plan is consistent with the optimization results output by S4;

[0150] (2) Train operation instructions: including dispatching commands for overtaking, waiting and avoiding, and adjusting the order of operation, wherein the dispatching commands must satisfy the set of operating permission boundary parameters defined by S3. ;

[0151] (3) Grouping operation trigger instruction: when When =1, a scheduling trigger request for grouping / degrouping operation is sent to the train control system. The train control system verifies the request based on the communication status and safety conditions. The dispatcher can confirm, modify, or reject the generated operation diagram adjustment plan.

[0152] S53, the CTC system synchronously transmits the operating permission boundary parameters to the train control system or RBC, thereby granting permission for operation and issuing control commands; wherein, the operating permission boundary parameters include:

[0153] (1) Mobility grant boundary parameters: Mobility grant distance reconstructed from S3 As recommended parameters for the operational permission boundary, these parameters are synchronized to the train control system, allowing the system to independently verify and execute them based on existing safety control logic, ensuring that train operation does not exceed the dynamic safety boundary.

[0154] (2) Minimum tracking interval and interval constraint parameters: The dynamic tracking interval With the minimum time of the interval As an operational control constraint, it is used by the train control system for speed monitoring and authorization calculation;

[0155] (3) Grouping control condition verification parameters: including communication link stability Safety margin The parameters and their threshold conditions are used to determine whether the train formation control command meets the execution conditions. After receiving the above parameters, the train control system completes the safety verification and issuance of the train traction, braking and formation control commands.

[0156] S54, Execution feedback and closed-loop correction, including: During the execution of the running instructions, the system collects feedback information in real time. The feedback information is used as a new running state input and sent back to S1 to trigger the update of the virtual grouping state and a new round of running trend prediction and optimization calculation, thereby forming a closed-loop scheduling control mechanism of "state perception-prediction-optimization-execution-feedback".

[0157] Preferably, the feedback information includes: the actual operating status of the train, including position, speed and acceleration; the actual arrival and departure time and the deviation from the execution; changes in communication status and control response time; and the execution result of the train formation operation.

[0158] A second aspect of the present invention provides an adaptive adjustment system for train timetables oriented towards virtual train formations, for implementing the method of the first aspect, comprising:

[0159] The data monitoring and acquisition module is used to monitor and acquire the raw status data of train operation in real time, and calculate the operating status parameters of the virtual train formation based on the raw status data;

[0160] The operation trend prediction module is used to form virtual group operation characteristic constraints based on the operation status parameters of the virtual group, and predict the operation trend based on the virtual group operation characteristic constraints to form operation trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operation permission boundaries and operation diagram adjustment;

[0161] The runtime permission boundary reconstruction module is used to dynamically reconstruct the runtime permission boundary based on the grouping state, thereby reconstructing the runtime constraints by predicting and real-time states and outputting the runtime permission boundary parameter set, so that the runtime graph adjustment has "adaptiveness".

[0162] The multi-objective adaptive optimization adjustment module is used to perform multi-objective adaptive optimization adjustment of the virtual train timetable using a multi-objective optimized timetable adaptive adjustment algorithm. This adjustment uses the set of operational permission boundary parameters output by S3 as constraints and combines them with the operational trend prediction results from S2 to generate an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation. The multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods. Those skilled in the art can implement the corresponding solution process based on the above objective function and constraints.

[0163] The collaborative control module is used to generate running instructions based on the optimized running graph adjustment results, and to perform collaborative control with CTC / RBC based on the running instructions.

[0164] A third aspect of the present invention provides an electronic device including a processor and a memory, the memory storing a plurality of instructions, the processor being configured to read the instructions and execute the method as described in the first aspect.

[0165] A fourth aspect of the present invention provides a computer-readable storage medium storing a plurality of instructions which can be read by a processor and executed as described in the first aspect.

[0166] The beneficial effects of the method and system of the present invention are as follows:

[0167] This invention has strong dynamic adjustment capabilities, good predictability, high safety margin, adaptability to virtual train formation scenarios, and close collaboration with the train control system. At the same time, it can improve the real-time performance and rationality of train timetable adjustments under virtual train formation operation conditions, enhance the dispatching system's adaptability to complex operating states, and provide technical support for the safe and efficient operation of virtual train formations. Attached Figure Description

[0168] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of the present invention, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0169] Figure 1 This is a flowchart of a train timetable adaptive adjustment method for virtual train formations provided according to an embodiment of the present invention;

[0170] Figure 2 This is a schematic diagram illustrating the coupling between operational trend prediction and virtual grouping characteristics according to an embodiment of the present invention;

[0171] Figure 3 This is a diagram illustrating the relationship between dynamic runtime license boundary construction and MA verification according to an embodiment of the present invention.

[0172] Figure 4 This is a schematic diagram illustrating virtual marshalling overtaking and unmarshalling under station resource constraints according to an embodiment of the present invention;

[0173] Figure 5 This is a structural diagram of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0174] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0175] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0176] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0177] like Figure 1-4 As shown, a first aspect of the present invention is to provide an adaptive adjustment method for train timetables for virtual train formations, comprising the following steps:

[0178] S1, monitor and collect raw status data of train operation in real time, and calculate the operating status parameters of virtual train formation based on the raw status data;

[0179] In a preferred embodiment, S1 includes:

[0180] S11, Collect raw status data of train operation, wherein the raw status data includes single-vehicle operation parameters, group train status and communication quality data;

[0181] In this embodiment, raw status data of train operation is collected in real time through interfaces with RBC, onboard system (ATO / ATP), etc., including:

[0182] (1) Single-vehicle operating parameters, including: train exist The position of the front of the car at that moment Rear of the car Longitudinal running speed Longitudinal acceleration and minimum emergency braking acceleration wait.

[0183] (2) Group train status, including: relative distance between trains Relative velocity wait.

[0184] (3) Communication quality data, including: vehicle-to-vehicle or vehicle-to-ground communication links. Communication strength at any given moment Packet loss rate Communication delay wait.

[0185] S12, Calculate the relative distance and speed difference between trains based on the original state data;

[0186] In this embodiment, for the following vehicle For example, it is related to the car in front. relative distance With speed difference As shown in equations (1) and (2) respectively:

[0187] (1);

[0188] (2);

[0189] Subscript Always represents the currently controlled train (following train), subscript It always represents the preceding vehicle associated with it;

[0190] S13, Evaluating link stability based on normalization function The link stability This forms the basis for whether virtual grouping can be maintained.

[0191] In this embodiment, the normalization function is shown in equations (3) and (4) below:

[0192] (3);

[0193] (4);

[0194] in, This represents the communication signal strength; a higher value indicates a more stable link. and These represent the upper and lower limits of the signal strength, respectively. Normalized communication signal strength; , and The three weighting coefficients are three non-negative fixed constants that do not change with time. change; for The probability of link error, packet loss, or failure at any given moment; for The communication link delay, transmission delay, or response delay at any given moment;

[0195] S14, Calculate the dynamic safety distance, including:

[0196] (1) Calculate the emergency braking distance As shown in equation (5):

[0197] (5);

[0198] in, For minimum emergency braking acceleration, This refers to the longitudinal running speed;

[0199] (2) During virtual train formation operation, due to the speed difference between the trains in front and behind, there may be a risk of longitudinal impact during emergency braking; therefore, a speed difference impact compensation distance is introduced. As shown in equation (6):

[0200] (6);

[0201] in, The impact compensation coefficient is used to reflect the longitudinal dynamic uncertainty in virtual train formation operation. Its value can be set according to the train type, train formation operation level or historical operation data.

[0202] (3) Considering that virtual train operation depends on vehicle-to-vehicle and vehicle-to-ground communication, communication delay and control response time may affect the safe distance. Therefore, a distance compensation method for calculating communication and control delay is introduced. The calculation formula is shown in equation (7):

[0203] (7);

[0204] Among them, the overall system delay time Due to communication delay Controlling computation delay and execution response latency Together they constitute the whole, and their calculation formula is shown in equation (8):

[0205] (8);

[0206] in, The data transmission delay used to characterize vehicle-to-vehicle and vehicle-to-ground communication is calculated using the formula shown in equation (9):

[0207] (9);

[0208] in, and They represent the first The sending and receiving times of the next communication data packet To count the number of data packets within the window;

[0209] The calculation formula is used to characterize the calculation and processing time of the scheduling system or train control system for the running instructions, as shown in equation (10):

[0210] (10);

[0211] in, This indicates the time when the scheduling system or train control system receives the status data. Indicates the time when the corresponding operation control command was generated and completed;

[0212] The formula used to characterize the train-end response time is shown in equation (11):

[0213] (11);

[0214] in, This indicates the time when the train receives the control command. Indicates the time at which the train begins to perform traction, braking, or train formation control actions;

[0215] S14, Calculate dynamic safety distance As shown in equation (12):

[0216] (12);

[0217] The dynamic safety distance Unlike traditional safety distances based on fixed braking models, this approach introduces communication and control delay compensation terms, enabling safety constraints to dynamically change with the virtual formation's operational status.

[0218] S15, based on the dynamic safety distance Calculate the safety margin of virtual groups The virtual grouping security margin Used to indicate whether the virtual train formation safety conditions are violated, its physical meaning is the difference between the current available safe distance between trains and the required minimum safe distance, calculated as shown in equation (13):

[0219] (13);

[0220] like This indicates a security instability and requires immediate triggering of S3 to tighten the permission boundaries;

[0221] S16, based on the virtual grouping security margin Constructing the state vector of the virtual group ;

[0222] In this embodiment, at time The state vector of the virtual group As shown in equation (14):

[0223] (14).

[0224] S2, based on the operating status parameters of the virtual group, virtual group operating characteristic constraints are formed, and based on the virtual group operating characteristic constraints, operating trends are predicted to form operating trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operating permission boundaries and adjustment of the operating diagram;

[0225] In a preferred embodiment, S2 includes:

[0226] S21, Constructing predictive input features, including: to meet the requirements of operational trend prediction, constructing the state vectors of the virtual grouping at consecutive time points as time series input, at time... Based on the state vector of the virtual group Forming an input state feature sequence for trend prediction. As shown in equation (15):

[0227] (15);

[0228] in, The state sampling time interval, This represents the length of the time series.

[0229] S22, considering the impact of communication link state fluctuations on the stability of train cooperative operation during virtual train formation operation, a communication uncertainty constraint mechanism is introduced in the operation trend prediction process, thereby forming a weighted state sequence under communication uncertainty constraints, including: based on communication link stability For the input state feature sequence Different weights are assigned to historical states at different times to reduce the impact of communication jitter and latency fluctuations on the prediction results of operational trends; when the communication link stability When the weighting is reduced, the influence weight of the long-term historical state on the prediction results is reduced, thereby enhancing the sensitivity of the prediction model to the current operating state.

[0230] S23, Model the nonlinear operating characteristics caused by ungrouping or regrouping the virtual train formation based on acceleration change characteristic quantities; wherein, the acceleration change characteristic quantities are used to reflect the impact of ungrouping or regrouping operations on the evolution of train operating state, and serve as one of the important input features of the prediction model;

[0231] In this embodiment, during the virtual train formation disassembly or reassembly process, the train traction and braking control strategies switch, and the train's longitudinal acceleration exhibits obvious nonlinear variation characteristics. To accurately characterize this operating characteristic, an acceleration variation characteristic quantity is introduced into the operating trend prediction process, as shown in equation (16):

[0232] (16);

[0233] in, The time step for running the trend prediction model, This refers to longitudinal acceleration.

[0234] S24, Construct a running trend prediction model, including: based on the input state feature sequence A time-series prediction model with a gating structure is constructed as an operation trend prediction model. The operation trend prediction model is used to model the time-series evolution relationship of the virtual train formation operation state. The gating structure is used to comprehensively consider the relative operation state changes between trains in the virtual train formation, the impact of communication link stability on the effectiveness of historical information, and the disturbance effect of nonlinear acceleration changes on the operation trend during the de-forming and re-forming process during the state update process.

[0235] In this embodiment, a gating mechanism is used to jointly model the long-term dependencies and short-term mutation characteristics of the virtual group's operating state. The time-series prediction model can be implemented using a rule-based prediction model, a statistical learning model, or a neural network model with a gating structure; this invention does not limit the specific implementation method.

[0236] S25, Output the virtual grouping operation trend prediction result based on the operation trend prediction model. ;

[0237] In this embodiment, the above prediction model outputs the prediction result of the running trend, as shown in equation (17):

[0238] (17);

[0239] in, For train At the station The predicted arrival time; For train At the station Predicted departure times; For train In the interval Predicted runtime; Predict the probability of conflict during virtual grouping operations; A predictive score for the stability of virtual group operation.

[0240] The predicted operational trend is not used for result evaluation, but rather serves as a direct input for calculating the dynamic operational permission boundary parameters. Without this prediction, it is impossible to construct a dynamic tracking interval that meets the virtual grouping operation conditions. With the minimum time of the interval .

[0241] Among them, the virtual grouping operation conflict prediction probability and virtual grouping stability score It can be obtained through historical operational data statistical models, rule models or machine learning models, and this invention does not limit its specific calculation method.

[0242] The above prediction results are used for subsequent dynamic reconstruction of operating permission boundaries and adaptive adjustment of train timetables, and are not directly used as the basis for real-time safety control of the train control system.

[0243] S3 dynamically reconstructs the operational permission boundary based on the grouping state, thereby predicting and reconstructing operational constraints in real time and outputting the operational permission boundary parameter set, making the operation diagram adjustment "adaptive".

[0244] In a preferred embodiment, S3 includes:

[0245] S31, the parameter set of the operation permission boundary that constitutes the operation permission boundary parameter system is defined. Under the virtual grouping operation condition, the parameter set of the operation permission boundary consists of the following key parameters, as shown in equation (18):

[0246] (18);

[0247] in, Indicates the dynamic tracking interval; Indicates the minimum time minute for interval operation; Indicates the permitted operating distance for the train; This represents the set of executable overrun time windows; This indicates the set of sections or stations that can be grouped / degrouped; the above parameters together constitute the train's schedule. Permissible operating boundaries;

[0248] S32, regarding the dynamic tracking interval Perform multi-factor reconstruction, including: the virtual grouping safety margin obtained based on S15. The virtual grouping operation conflict prediction probability obtained by S2 and virtual grouping stability score The minimum safe tracking interval is dynamically reconstructed, as shown in equation (19):

[0249] (19);

[0250] in: The baseline tracking interval is set under ideal operating conditions. For virtual grouping safety margin; Predict the probability of conflict during virtual grouping operations; Assess the stability of virtual groupings; , , The weight coefficients are non-negative and The three weights correspond to three types of operational constraints that cannot be ignored in parallel: safety instability triggering, conflict risk amplification, and group stability degradation. They adopt a linear superposition form to ensure that the tracking interval remains monotonically constant when any risk factor deteriorates.

[0251] S33, Based on the physical length of the interval and the predicted interval running time, calculate the minimum time division of the interval. Dynamic correction is performed, as shown in equation (20):

[0252] (20);

[0253] in, For the time interval of the graph, The risk correction coefficient for the minimum time division of the interval is used to reflect the impact of communication uncertainty, changes in group stability, and operational conflict risk on the interval's operational safety margin during virtual grouping operation, as shown in equation (21):

[0254] (twenty one);

[0255] in, , and The weight coefficients are non-negative and ;

[0256] S34, the available operating distance of the train The recommended parameters for the operational clearance boundary are synchronized to the train control system, which then determines the available operational clearance distance for the train based on the dynamic safety distance and existing safety control logic. Conduct security checks and implement them;

[0257] The train's available operating permit distance The suggested operating permission boundary value is calculated by the scheduling layer based on the virtual grouping operation status and is used to constrain the operation diagram adjustment results. Its final execution still needs to be verified by the train control system for safety.

[0258] In this embodiment, at time The available operating distance of the train calculated by the scheduling layer The following constraint relationship is satisfied:

[0259] (twenty two);

[0260] in, The dynamic safety distance calculated in step S14; The maximum authorized position is limited by track conditions or the capabilities of the train control system;

[0261] The train's available operating permit distance It is not a real-time movement authorization for the train control system, but rather an upper bound condition used to constrain the feasible solution space for adjusting the operation diagram. By excluding adjustment schemes that violate the dynamic safety distance in advance at the scheduling layer, invalid or unexecutable operation diagram results are prevented from being sent to the train control system.

[0262] The aforementioned security verification process is used to ensure that the recommended operating permission boundary values ​​generated by the scheduling layer are always within the range allowed by the security protection capabilities of the train control system, thereby ensuring the consistency of scheduling instructions and train control security protection in terms of data and logic.

[0263] S35, Filter Override and Group Resource Availability Window, including:

[0264] (1) Filter the set of overrun time windows As shown in equation (23):

[0265] (twenty three);

[0266] in: Indicates at the station At this point, the predicted time margin for train overtaking or decoupling operations under operating conditions is used to assess whether station operations have a feasible time window. This indicates a time conflict in station operations, making overtaking or decoupling impossible. The calculation formula is shown in equation (24):

[0267] (twenty four);

[0268] in, Trains predicted for S2 Arrival at the station The moment; Trains predicted for S2 Leaving the station The moment; To address prediction errors, communication jitter, and job execution uncertainties, a time safety redundancy is introduced, calculated as shown in equation (23):

[0269] (twenty three);

[0270] in, The average time for station route arrangement and scheduling is calculated. The average transmission delay between scheduling instructions and train control commands can be reused from S14. ; The physical execution time for on-site route planning, turnout switching to signal opening; The fixed safety redundancy time configured for the system is preferably 5 to 30 seconds;

[0271] This is a track capacity indicator used to indicate whether station s has the capacity to handle the current virtual marshalling operation within the prediction time window. The judgment condition is shown in equation (24):

[0272] (twenty four);

[0273] The track length, track occupancy status, and route availability information are provided by the CTC station basic database and real-time route management module.

[0274] The time safety margin of the station is calculated based on the operational trend prediction results. In conjunction with the system's reserved safety time and station track capacity assessment indicators This enables unified constraint judgment on the feasibility of station overtaking and virtual marshalling / re-marshalling.

[0275] (2) Establish a set of grouping / degrouping sections As shown in equation (25):

[0276] (25);

[0277] in: The communication link stability calculated for S13; This is the threshold for link stability; This refers to the length of available tracks at the station. This refers to the total length of the virtual train formation.

[0278] S36, Output runtime license boundaries, including: the runtime license boundary parameter set obtained from dynamic reconstruction. The output is sent to S4 as a dynamic constraint condition for the multi-objective adaptive adjustment model of the running graph, ensuring that the generated running graph adjustment scheme meets the requirements of virtual grouping operation in terms of security, executability and communication reliability.

[0279] S4, using a multi-objective optimization adaptive adjustment algorithm for the runtime graph, takes the runtime permission boundary parameter set output by S3 as an example. As constraints, and in conjunction with the predicted operational trends of S2, the virtual train timetable is subjected to multi-objective adaptive optimization adjustment to generate an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation. The multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods. Those skilled in the art can implement the corresponding solution process based on the above objective function and constraints.

[0280] In a preferred embodiment, S4 includes:

[0281] S41, optimize the input and define the decision variables; including:

[0282] (1) Optimize input, including:

[0283] The dynamic runtime permission boundary is optimized as shown in equation (26):

[0284] (26);

[0285] The optimized trend prediction output is shown in equation (27):

[0286] (27);

[0287] (2) Define decision variables, including:

[0288] Define the set of decision variables for running graph adjustment As shown in equation (28):

[0289] (28);

[0290] in, Indicates train At the station Adjustment amount for arrival and departure times; Indicates train At the station Stock market selection decisions; This represents the virtual group switching decision variable, where 1 indicates performing decompilation / recompilation;

[0291] S42, construct the comprehensive multi-objective optimization objective function, as shown in equation (29):

[0292] (29);

[0293] in, , , , These are the weighting coefficients, and =1;

[0294] For punctual targets, the calculation formula is shown in equation (30):

[0295] (30);

[0296] in, , , and The diagram shows the arrival and departure times, where... and These represent the adjusted trains. At the station Arrival time;

[0297] For the purpose of smooth operation or energy consumption, the adjustment range of operating time is used as a proxy indicator of energy consumption and comfort. The calculation formula is shown in equation (31):

[0298] (31);

[0299] in, and They represent trains The start and end times within the current optimization window; Indicates train At any moment The longitudinal acceleration of the movement; This represents the rate of change of the train's longitudinal acceleration over time. and This is the smoothness weighting coefficient; Used to constrain operating efficiency and energy consumption. Used to constrain operational comfort and longitudinal stability of train formation; the operational smoothness target or energy consumption proxy target is only used as an evaluation index for timetable adjustment and does not directly participate in train traction or braking control;

[0300] For the grouping stability target, the calculation is shown in equation (32):

[0301] (32);

[0302] To implement a conflict risk penalty, which guides the system to prioritize the operating scheme with lower conflict risk when multiple solutions are feasible, the calculation is shown in equation (33):

[0303] (33);

[0304] S43, define the constraints, including minimum time division constraint, dynamic tracking interval constraint, movement authorization constraint, and station resources and marshalling operation constraint;

[0305] The minimum time-division constraint of the interval is shown in equation (34):

[0306] (34);

[0307] When the optimizer finds that it cannot... When a feasible solution is found within the constraints, the system will trigger a priority inversion mechanism, prioritizing the sacrifice of the punctuality objective. To ensure a safety margin ;

[0308] The dynamic tracking interval constraint is shown in equation (35):

[0309] (35);

[0310] The mobility authorization constraint is shown in equation (36):

[0311] (36);

[0312] The station resources and marshalling operation constraints are shown in equations (37) and (38):

[0313] (37);

[0314] (38);

[0315] It should be noted that the aforementioned movement authorization constraints are only used for determining the feasibility of the operational diagram and filtering the optimization solution space, and do not participate in the train control system's real-time... The secure computation process.

[0316] S44, based on heuristic search and a rolling time-domain multi-objective optimization strategy, performs adaptive solution and rolling update, including:

[0317] (1) Initialization: based on the current state Compared with the prediction results Initial value;

[0318] (2) Window scrolling: Set up scrolling optimization for windows, in the window [ Internally solve for the optimal adjustment sequence;

[0319] (3) Conflict detection and resolution: If a conflict is detected Automatically increase the weight of safety and risk objectives;

[0320] (4) Scheme output: Output the executable running graph adjustment scheme at the current time.

[0321] (5) When a feasible solution that satisfies all constraints is not obtained within the current rolling optimization window, the system automatically reverts to the most recently executable running graph state and increases the weight of safety and risk-related objectives to solve the problem again.

[0322] S45, Output the optimized runtime chart adjustment results. And pass it to S5 for execution. The optimized runtime graph adjustment result is shown in equation (39):

[0323] (39);

[0324] in, This indicates the train's arrival time after adjustments at each station. This indicates the train's departure time after adjustments at each station. This indicates the decision-making process for train track occupancy at the station, the corresponding route usage window, and the range of the occupied section; Indicates a virtual grouping status switching command;

[0325] S5, Adjust the results based on the optimized running graph. Generate execution instructions and perform coordinated control with CTC / RBC based on the execution instructions.

[0326] Adjust the optimized running graph output by S4. Forming a grouping control strategy and converting it into operating instructions that can be executed by the scheduling system and the train control system, thereby achieving closed-loop collaboration between the scheduling layer and the control layer;

[0327] In a preferred embodiment, S5 includes:

[0328] S51, parse the run graph adjustment results and form an instruction parameter mapping, including: based on the optimized run graph adjustment results obtained in step S4. The results of the runtime chart adjustment are analyzed in a structured manner, and the following three types of instruction parameters are generated as the input basis for the generation of runtime instructions:

[0329] (1) Time-related parameters, including: the adjusted arrival times of trains at each station. Adjusted departure times of trains at various stations ;

[0330] (2) Resource parameters, including: track occupancy decisions for trains at stations. And the corresponding route usage window and occupied section range;

[0331] (3) Grouping control parameters, including: virtual grouping status switching instructions It is used to instruct trains to form, detach, or maintain the current formation status.

[0332] S52, the centralized scheduling system generates scheduling layer operation instructions for the operation organization layer based on the parsing of the instruction parameters. Multiple scheduling layer operation instructions form a scheduling instruction set, including:

[0333] (1) Train timetable adjustment plan: used to update the arrival and departure times, overtaking relationships and track utilization plans in the train timetable, so that the scheduling plan is consistent with the optimization results output for S4.

[0334] (2) Train operation instructions: including dispatching commands for overtaking, waiting and avoiding, and adjusting the order of operation, wherein the dispatching commands must satisfy the set of operating permission boundary parameters defined by S3. .

[0335] (3) Grouping operation trigger command:

[0336] when When =1, a scheduling trigger request for grouping / degrouping operation is sent to the train control system. The train control system verifies the request based on the communication status and safety conditions. The dispatcher can confirm, modify, or reject the generated operation diagram adjustment plan.

[0337] S53, the CTC system synchronously transmits the operating permission boundary parameters to the train control system or RBC, thereby granting permission for operation and issuing control commands; wherein, the operating permission boundary parameters include:

[0338] (1) Mobility grant boundary parameters: Mobility grant distance reconstructed from S3 As recommended parameters for the operational permission boundary, these parameters are synchronized to the train control system, allowing the system to independently verify and execute them based on existing safety control logic, ensuring that train operation does not exceed the dynamic safety boundary.

[0339] (2) Minimum tracking interval and interval constraint parameters: The dynamic tracking interval With the minimum time of the interval As an operational control constraint, it is used by the train control system for speed monitoring and authorization calculation.

[0340] (3) Grouping control condition verification parameters: including communication link stability Safety margin The parameters and their threshold conditions are used to determine whether the train formation control command meets the execution conditions. After receiving the above parameters, the train control system completes the safety verification and issuance of the train traction, braking and formation control commands.

[0341] The aforementioned operational permission boundary parameters serve only as suggestions for operational organization constraints at the scheduling layer and do not directly replace the safety calculation logic of the train control system for movement authorization, speed monitoring, and braking control.

[0342] S54, Execution Feedback and Closed-Loop Correction, includes: During the execution of running instructions, the system collects feedback information in real time. This feedback information serves as a new operating state input and is transmitted back to S1, triggering the update of the virtual grouping state and a new round of operating trend prediction and optimization calculation, thereby forming a closed-loop scheduling control mechanism of "state perception-prediction-optimization-execution-feedback"; wherein, the feedback information includes:

[0343] (1) Actual operating status of the train (position, speed, acceleration);

[0344] (2) Deviation between actual arrival and departure times and execution time;

[0345] (3) Communication status changes and control response time;

[0346] (4) Results of grouping operation.

[0347] Through the aforementioned scheduling layer-train control layer collaborative mechanism, the operation diagram adjustment under virtual grouping operation conditions has an integrated technical path of prediction-driven, boundary constraint, and closed-loop execution.

[0348] A second aspect of the present invention provides an adaptive adjustment system for train timetables oriented towards virtual train formations, for implementing the method of the first aspect, comprising:

[0349] The data monitoring and acquisition module is used to monitor and acquire the raw status data of train operation in real time, and calculate the operating status parameters of the virtual train formation based on the raw status data;

[0350] The operation trend prediction module is used to form virtual group operation characteristic constraints based on the operation status parameters of the virtual group, and predict the operation trend based on the virtual group operation characteristic constraints to form operation trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operation permission boundaries and operation diagram adjustment;

[0351] The runtime permission boundary reconstruction module is used to dynamically reconstruct the runtime permission boundary based on the grouping state, thereby reconstructing the runtime constraints by predicting and real-time states and outputting the runtime permission boundary parameter set, so that the runtime graph adjustment has "adaptiveness".

[0352] The multi-objective adaptive optimization adjustment module is used to perform multi-objective adaptive optimization adjustment of the virtual train timetable using a multi-objective optimized timetable adaptive adjustment algorithm. This adjustment uses the set of operational permission boundary parameters output by S3 as constraints and combines them with the operational trend prediction results from S2 to generate an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation. The multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods. Those skilled in the art can implement the corresponding solution process based on the above objective function and constraints.

[0353] The collaborative control module is used to generate running instructions based on the optimized running graph adjustment results, and to perform collaborative control with CTC / RBC based on the running instructions.

[0354] The present invention also provides a memory that stores multiple instructions for implementing the method as described in Embodiment 1.

[0355] like Figure 5 As shown, the present invention also provides an electronic device, including a processor 301 and a memory 302 connected to the processor 301. The memory 302 stores a plurality of instructions, which can be loaded and executed by the processor to enable the processor to perform the method as described in Embodiment 1.

[0356] Application Example: An Example of an Adaptive Adjustment Method for Run Graphs in Virtual Grouping Run Scenarios

[0357] This application example uses a railway line with virtual train formation operation capability as the application scenario. Multiple trains with train-to-train communication capability operate on the line simultaneously, and the dispatching system and train control system achieve data interaction through a standard interface.

[0358] The adaptive adjustment method for the running graph described in this embodiment includes the following steps:

[0359] (1) Virtual grouping status acquisition and real-time monitoring

[0360] In this embodiment, the system periodically collects train operation status data, including train position, speed, acceleration, braking performance parameters, and current virtual formation structure information, through the CTC system interface with the train control system; at the same time, it collects communication quality data such as signal strength, communication latency, and packet loss rate of train-to-train and train-to-ground communication.

[0361] Based on the collected data, the system calculates the relative distance and relative speed between trains in the virtual train formation in real time, and evaluates the stability of the communication link by normalizing the communication indicators.

[0362] Based on this, the system combines train braking performance, speed difference, and communication and control delay to dynamically calculate the safe distance required for virtual train formation operation, and further obtains the formation safety margin to determine whether the current formation operation status is within a safe and controllable range.

[0363] (2) Operational trend prediction

[0364] In this embodiment, the system constructs the virtual grouping status acquired at continuous time intervals into time series input features and inputs them into the running trend prediction model.

[0365] The prediction model adopts a time-series prediction structure with a gating mechanism. By updating and filtering historical operating states, it outputs the prediction results of the operating trend in the future prediction time domain, including: train arrival and departure time deviation prediction, section running time change trend, probability of running conflict, and virtual train formation stability score.

[0366] The prediction results serve as an important basis for subsequent operation permission boundary reconstruction and operation diagram adjustment.

[0367] (3) Dynamic reconfiguration of runtime permission boundaries based on grouping status

[0368] In this embodiment, the system dynamically reconstructs the virtual grouping operation permission boundary based on real-time grouping safety margin and predicted operation trend.

[0369] Specifically, the system adaptively corrects the minimum tracking interval between trains based on the predicted train formation stability score and conflict probability; at the same time, it dynamically adjusts the minimum running time of the interval based on the predicted interval running speed.

[0370] In addition, the system performs a safety check on the moving authorization endpoint based on the position of the preceding vehicle and the dynamic safety distance to ensure that the scheduling adjustment results do not exceed the physical safety boundary of the train control system.

[0371] The system also dynamically filters station or section resource sets that can be used for overtaking and marshalling operations based on station capacity, marshalling length, and communication link stability.

[0372] Through the above processing, a set of dynamic operational permission boundary parameters for the current moment is formed, providing constraints for operational graph optimization.

[0373] (4) Adaptive adjustment of the running graph for multi-objective optimization

[0374] In this embodiment, the system uses the dynamic operating permission boundary output in step S3 as a constraint to construct a multi-objective optimization model and adaptively adjust the train timetable.

[0375] The optimization model uses train arrival and departure times, track selection, and virtual formation state switching as decision variables, and comprehensively considers multiple optimization objectives such as punctuality, smooth operation, energy efficiency, formation stability, and conflict risk.

[0376] The system adopts a rolling time-domain optimization method to search for the optimal operation graph adjustment scheme that satisfies the dynamic operation permission boundary constraints within a preset time window, thereby obtaining the optimal operation time solution and the corresponding operation organization decision.

[0377] (5) Execution instruction generation and coordinated control

[0378] In this embodiment, the system parses the optimal timetable adjustment result output in step S4 into executable scheduling instructions, and issues timetable adjustment instructions, overtaking organization instructions, and train formation control trigger instructions to relevant trains through the centralized scheduling system.

[0379] At the same time, the system will synchronously transmit the dynamically reconstructed movement authorization boundary, minimum tracking interval and interval operation constraints to the train control system, which will then complete the operation permission verification and control command issuance.

[0380] During train operation, the system acquires execution feedback information in real time and uses it as a new state input to trigger the next round of operation trend prediction and operation diagram optimization, thereby forming a closed-loop virtual train formation operation scheduling and control process.

[0381] (6) Implementation results

[0382] Through the above embodiments, the present invention realizes real-time perception, predictive adjustment and dynamic reconstruction of train timetable in virtual train formation operation scenarios, which can effectively improve operational safety, punctuality and line capacity utilization efficiency, and is suitable for high-density transportation and complex operation organization scenarios.

[0383] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for adaptive adjustment of train timetables for virtual train formations, characterized in that, include: S1, monitor and collect raw status data of train operation in real time, and calculate the operating status parameters of virtual train formation based on the raw status data; S2, based on the operating status parameters of the virtual group, virtual group operating characteristic constraints are formed, and based on the virtual group operating characteristic constraints, operating trends are predicted to form operating trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operating permission boundaries and adjustment of the operating diagram; S3 dynamically reconstructs the operational permission boundary based on the grouping state, thereby predicting and reconstructing operational constraints in real time and outputting the operational permission boundary parameter set, making the operation diagram adjustment "adaptive". S4, using a multi-objective adaptive adjustment algorithm for the train schedule, with the set of operational permission boundary parameters output in S3 as constraints, and combined with the operational trend prediction results in S2, performs multi-objective adaptive optimization adjustment on the virtual train schedule, generating an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation as the optimized train schedule adjustment result; wherein, the multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods, and the corresponding solution process is implemented according to the objective function and constraints; S5, generate running instructions based on the optimized running diagram adjustment results, and perform coordinated control with CTC / RBC based on the running instructions.

2. The adaptive adjustment method for train timetables for virtual train formations according to claim 1, characterized in that, S1 includes: S11, Collect raw status data of train operation, wherein the raw status data includes single-vehicle operation parameters, group train status and communication quality data; Raw train operation status data is collected in real time through interfaces with the RBC and onboard systems, including: (1) Single-vehicle operating parameters, including: train exist The position of the front of the car at that moment Rear of the car Longitudinal running speed Longitudinal acceleration and minimum emergency braking acceleration ; (2) Group train status, including: relative distance between trains and relative velocity ; (3) Communication quality data, including: vehicle-to-vehicle or vehicle-to-ground communication links. Communication strength at any given moment Packet loss rate Communication delay ; S12, Calculate the relative distance and speed difference between trains based on the original state data; For the car behind For example, it is related to the car in front. relative distance With speed difference As shown in equations (1) and (2) respectively: （1）； （2）； S13, Evaluating link stability based on normalization function The normalization function is shown in equations (3) and (4) below: （3）； （4）； in, This represents the communication signal strength; a higher value indicates a more stable link. and These represent the upper and lower limits of the signal strength, respectively. Normalized communication signal strength; , and The three weighting coefficients are three non-negative fixed constants that do not change with time. change; for The probability of link error, packet loss, or failure at any given moment; for The communication link delay, transmission delay, or response delay at any given moment; S14, Calculate the dynamic safety distance, including: (1) Calculate the emergency braking distance As shown in equation (5): （5）； in, For minimum emergency braking acceleration, This refers to the longitudinal running speed; (2) Calculate the impact compensation distance for velocity difference As shown in equation (6): （6）； in, The impact compensation coefficient is used to reflect the longitudinal dynamic uncertainty in virtual train formation operation. Its value can be set according to the train type, train formation operation level or historical operation data. (3) Calculate the communication and control delay compensation distance The calculation formula is shown in equation (7): （7）； Among them, the overall system delay time Due to communication delay Controlling computation delay and execution response latency Together they constitute the whole, and their calculation formula is shown in equation (8): （8）； in, The data transmission delay used to characterize vehicle-to-vehicle and vehicle-to-ground communication is calculated using the formula shown in equation (9): （9）； in, and They represent the first The sending and receiving times of the next communication data packet To count the number of data packets within the window; The calculation formula is used to characterize the calculation and processing time of the scheduling system or train control system for the running instructions, as shown in equation (10): （10）； in, This indicates the time when the scheduling system or train control system receives the status data. Indicates the time when the corresponding operation control command was generated and completed; The formula used to characterize the train-end response time is shown in equation (11): （11）； in, This indicates the time when the train receives the control command. Indicates the time at which the train begins to perform traction, braking, or train formation control actions; S15, Calculate dynamic safety distance As shown in equation (12): （12）； The dynamic safety distance Unlike traditional safety distances based on fixed braking models; S16, based on the dynamic safety distance Calculate the safety margin of virtual grouping The calculation is shown in equation (13): （13）； like This indicates a security instability and requires immediate triggering of S3 to tighten the permission boundaries; S17, based on the virtual grouping security margin Constructing the state vector of the virtual group At any moment The state vector of the virtual group As shown in equation (14): （14）。 3. The adaptive adjustment method for train timetables for virtual train formations according to claim 2, characterized in that, S2 includes: S21, Construct the predicted input features, including: at time... Based on the state vector of the virtual group Forming an input state feature sequence for trend prediction. As shown in equation (15): （15）； in, The state sampling time interval, The length of the time series; S22, a communication uncertainty constraint mechanism is introduced during the trend prediction process, thereby forming a weighted state sequence under communication uncertainty constraints, including: based on the stability of the communication link. For the input state feature sequence Different weights are assigned to historical states at different times to reduce the impact of communication jitter and latency fluctuations on the prediction results of operational trends; S23, Based on the acceleration change characteristic quantity, the nonlinear operating characteristics caused by the ungrouping or regrouping of the virtual group are modeled, and the acceleration change characteristic quantity is introduced in the operation trend prediction process, as shown in equation (16): （16）； in, The time step for running the trend prediction model, For longitudinal acceleration; S24, Construct a running trend prediction model, including: based on the input state feature sequence A time-series prediction model with a gating structure is constructed as an operation trend prediction model. The operation trend prediction model is used to model the time-series evolution relationship of the virtual train formation operation state. The gating structure is used to comprehensively consider the relative operation state changes between trains in the virtual train formation, the impact of communication link stability on the effectiveness of historical information, and the disturbance effect of nonlinear acceleration changes on the operation trend during the de-forming and re-forming process during the state update process. S25, Output the virtual grouping operation trend prediction result based on the operation trend prediction model. As shown in equation (17): （17）； in, For train At the station The predicted arrival time; For train At the station Predicted departure times; For train In the interval Predicted runtime; Predict the probability of conflict during virtual grouping operations; A predictive score for the stability of virtual group operation.

4. The adaptive adjustment method for train timetables for virtual train formations according to claim 3, characterized in that, S3 includes: S31, the parameter set of the operation permission boundary that constitutes the operation permission boundary parameter system is defined. Under the virtual grouping operation condition, the parameter set of the operation permission boundary consists of the following key parameters, as shown in equation (18): （18）； in, Indicates the dynamic tracking interval; Indicates the minimum time minute for interval operation; Indicates the permitted operating distance for the train; This represents the set of executable overrun time windows; This indicates the set of sections or stations that can be grouped / degrouped; the above parameters together constitute the train's schedule. Permissible operating boundaries; S32, regarding the dynamic tracking interval Perform multi-factor reconstruction, including: the virtual grouping safety margin obtained based on S15. The virtual grouping operation conflict prediction probability obtained by S2 and virtual grouping stability score The minimum safe tracking interval is dynamically reconstructed, as shown in equation (19): （19）； in: The baseline tracking interval is set under ideal operating conditions. For virtual grouping safety margin; Predict the probability of conflict during virtual grouping operations; Assess the stability of virtual groupings; , , The weight coefficients are non-negative and ; S33, Based on the physical length of the interval and the predicted interval running time, calculate the minimum time division of the interval. Dynamic correction is performed, as shown in equation (20): （20）； in, For the time interval of the graph, The risk correction coefficient for the minimum time division of the interval is used to reflect the impact of communication uncertainty, changes in group stability, and operational conflict risk on the interval's operational safety margin during virtual grouping operation, as shown in equation (21): （21）； in, , and The weight coefficients are non-negative and ; S34, the available operating distance of the train The recommended parameters for the operational clearance boundary are synchronized to the train control system, which then determines the available operational clearance distance for the train based on the dynamic safety distance and existing safety control logic. Conduct security checks and implement them at all times; The available operating distance of the train calculated by the scheduling layer The following constraint relationship is satisfied: （22）； in, The dynamic safety distance calculated in step S14; The maximum authorized position is limited by track conditions or the capabilities of the train control system; S35, Filter Override and Group Resource Availability Window, including: (1) Filter the set of overrun time windows As shown in equation (23): （23）； in: Indicates at the station At this point, the predicted time margin for train overtaking or decoupling operations under operating conditions is used to assess whether station operations have a feasible time window. This indicates a time conflict in station operations, making overtaking or decoupling impossible. The calculation formula is shown in equation (24): （24）； in, Trains predicted for S2 Arrival at the station The moment; Trains predicted for S2 Leaving the station The moment; To address prediction errors, communication jitter, and job execution uncertainties, a time safety redundancy is introduced, calculated as shown in equation (25): （25）； in, The average time for station route arrangement and scheduling is calculated. The average transmission delay between scheduling instructions and train control commands can be reused from S14. ; The physical execution time for on-site route planning, turnout switching to signal opening; A fixed safety redundancy time configured for the system; This is a track capacity indicator used to indicate whether station s has the capacity to handle the current virtual marshalling operation within the prediction time window. The determination condition is shown in equation (26): （26）； The track length, track occupancy status, and route availability information of the station are provided by the CTC station basic database and real-time route management module; (2) Establish a set of grouping / degrouping sections As shown in equation (27): （27）； in: The communication link stability calculated for S13; This is the threshold for link stability. This refers to the length of available tracks at the station. This refers to the total length of the virtual train formation. S36, Output runtime license boundaries, including: the runtime license boundary parameter set obtained from dynamic reconstruction. The output is sent to S4 as a dynamic constraint condition for the multi-objective adaptive adjustment model of the running graph, ensuring that the generated running graph adjustment scheme meets the requirements of virtual grouping operation in terms of security, executability and communication reliability.

5. The adaptive adjustment method for train timetables for virtual train formations according to claim 4, characterized in that, S4 includes: S41, optimize the input and define the decision variables; including: (1) Optimize input, including: Optimize the dynamic runtime permission boundary, as shown in equation (28): （28）； The optimized trend prediction output is shown in equation (29): （29）； (2) Define decision variables, including: Define the set of decision variables for running graph adjustment As shown in equation (30): （30）； in, Indicates train At the station Adjustment amount for arrival and departure times; Indicates train At the station Stock market selection decisions; This represents the virtual group switching decision variable, where 1 indicates performing decompilation / recompilation; S42, construct the comprehensive multi-objective optimization objective function, as shown in equation (31): （31）； in, , , , These are the weighting coefficients, and =1; For punctual targets, the calculation formula is shown in equation (32): （32）； in, , , and The diagram shows the arrival and departure times, where... and These represent the adjusted trains. At the station Arrival time; To achieve the smoothness target or energy consumption proxy target, the operating time adjustment range is used as a proxy indicator for energy consumption and comfort. The calculation formula is shown in equation (33): （33）； in, and They represent trains The start and end times within the current optimization window; Indicates train At any moment The longitudinal acceleration of the movement; This represents the rate of change of the train's longitudinal acceleration over time. and This is the smoothness weighting coefficient; Used to constrain operating efficiency and energy consumption. Used to constrain operational comfort and longitudinal stability of train formation; the operational smoothness target or energy consumption proxy target is only used as an evaluation index for timetable adjustment and does not directly participate in train traction or braking control; For the group stability target, the calculation is shown in equation (34): （34）； To implement a conflict risk penalty, which guides the system to prioritize the operating scheme with lower conflict risk when multiple solutions are feasible, the calculation is shown in equation (35): （35）； S43, define the constraints, including minimum time division constraints, dynamic tracking interval constraints, movement authorization constraints, and station resources and marshalling operation constraints; The minimum time-division constraint of the interval is shown in equation (36): （36）； When the optimizer finds that it cannot... When a feasible solution is found within the constraints, the system will trigger a priority inversion mechanism, prioritizing the sacrifice of the punctuality objective. To ensure a safety margin ; The dynamic tracking interval constraint is shown in equation (37): （37）； The mobility authorization constraints are as shown in equation (38): （38）； The station resources and marshalling operation constraints are shown in equations (39) and (40): （39）； （40）； S44, based on heuristic search and a rolling time-domain multi-objective optimization strategy, performs adaptive solution and rolling update, including: (1) Initialization: based on the current state Compared with the prediction results Initial value; (2) Window scrolling: Set up scrolling optimization for windows, in the window [ Internally solve for the optimal adjustment sequence; (3) Conflict detection and resolution: If a conflict is detected Automatically increase the weight of safety and risk objectives; (4) Scheme output: Output the executable runtime adjustment scheme at the current time; (5) When a feasible solution that satisfies all constraints cannot be obtained within the current rolling optimization window, the system automatically reverts to the most recently executable running graph state and increases the weight of safety and risk-related objectives to solve the problem again. S45, Output the optimized runtime chart adjustment results. And pass it to S5 for execution. The optimized runtime graph adjustment result is shown in equation (41): （41）； in, This indicates the train's arrival time after adjustments at each station. This indicates the train's departure time after adjustments at each station. This indicates the decision-making process for train track occupancy at the station, the corresponding route usage window, and the range of the occupied section; This indicates a command to switch virtual group status.

6. The adaptive adjustment method for train timetables for virtual train formations according to claim 5, characterized in that, S5 includes: S51, parse the run graph adjustment results and form an instruction parameter mapping, including: the optimized run graph adjustment results obtained in S4. The results of the runtime chart adjustment are analyzed in a structured manner, and the following three types of instruction parameters are generated as the input basis for the generation of runtime instructions: (1) Time-related parameters, including: the adjusted arrival times of trains at each station. Adjusted departure times of trains at various stations ; (2) Resource parameters, including: track occupancy decisions for trains at stations. And the corresponding route usage window and occupied section range; (3) Grouping control parameters, including: virtual grouping status switching instructions This is used to instruct trains to form, detach, or maintain their current formation status. S52, the centralized scheduling system generates scheduling layer operation instructions for the operation organization layer based on the parsing of the instruction parameters. Multiple scheduling layer operation instructions form a scheduling instruction set, including: (1) Train timetable adjustment plan: used to update the arrival and departure times, overtaking relationships and track utilization plans in the train timetable, so that the scheduling plan is consistent with the optimization results output for S4; (2) Train operation instructions: including dispatching commands for overtaking, waiting and avoiding, and adjusting the order of operation. The dispatching commands must satisfy the set of operating permission boundary parameters defined by S3. ; (3) Grouping operation trigger instruction: when When =1, a scheduling trigger request for grouping / degrouping operation is sent to the train control system. The train control system verifies the request based on the communication status and safety conditions. The dispatcher can confirm, modify, or reject the generated operation diagram adjustment plan. S53, the CTC system synchronously transmits the operating permission boundary parameters to the train control system or RBC, thereby granting permission for operation and issuing control commands; wherein, the operating permission boundary parameters include: (1) Mobility grant boundary parameters: Mobility grant distance reconstructed from S3 As a suggested parameter for the operating permission boundary, it is synchronized to the train control system, so that the train control system can independently verify and execute it based on the existing safety control logic to ensure that the train operation does not exceed the dynamic safety boundary; (2) Minimum tracking interval and interval constraint parameters: The dynamic tracking interval With the minimum time of the interval As an operational control constraint, it is used by the train control system for speed monitoring and authorization calculation; (3) Grouping control condition verification parameters: including communication link stability Safety and security The parameters and their threshold conditions are used to determine whether the train formation control command meets the execution conditions. After receiving the above parameters, the train control system completes the safety verification and issuance of the train traction, braking and formation control commands. S54, Execution feedback and closed-loop correction, including: During the execution of the running instructions, the system collects feedback information in real time. The feedback information is used as a new running state input and sent back to S1 to trigger the update of the virtual grouping state and a new round of running trend prediction and optimization calculation, thereby forming a closed-loop scheduling control mechanism of "state perception-prediction-optimization-execution-feedback".

7. The adaptive adjustment method for train timetables for virtual train formations according to claim 6, characterized in that, The feedback information includes: the actual operating status of the train, including position, speed and acceleration; the actual arrival and departure times and the deviation from the execution; changes in communication status and control response time; and the results of the train formation operation.

8. A train timetable adaptive adjustment system for virtual train formations, used to implement the method described in any one of claims 1-7, characterized in that, include: The data monitoring and acquisition module is used to monitor and acquire the raw status data of train operation in real time, and calculate the operating status parameters of the virtual train formation based on the raw status data; The operation trend prediction module is used to form virtual group operation characteristic constraints based on the operation status parameters of the virtual group, and predict the operation trend based on the virtual group operation characteristic constraints to form operation trend prediction results, thereby providing predictive information support for subsequent dynamic reconstruction of operation permission boundaries and operation diagram adjustment; The runtime permission boundary reconstruction module is used to dynamically reconstruct the runtime permission boundary based on the grouping state, thereby reconstructing the runtime constraints by predicting and real-time states and outputting the runtime permission boundary parameter set, so that the runtime graph adjustment has "adaptiveness". The multi-objective adaptive optimization adjustment module is used to perform multi-objective adaptive optimization adjustment of the virtual train timetable using a multi-objective optimized timetable adaptive adjustment algorithm. This adjustment uses the set of operational permission boundary parameters output by S3 as constraints and combines them with the operational trend prediction results from S2 to generate an adjustment scheme that meets the requirements of safety, punctuality, stability, and coordinated operation. The multi-objective adaptive optimization adjustment is solved using linear programming, integer programming, or heuristic optimization methods. Those skilled in the art can implement the corresponding solution process based on the above objective function and constraints. The collaborative control module is used to generate running instructions based on the optimized running graph adjustment results, and to perform collaborative control with CTC / RBC based on the running instructions.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing multiple instructions, and the processor being used to read the instructions and execute the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions, which can be read by a processor and executed as described in any one of claims 1-7.

Images