A group train tracking operation decision method and system considering communication delay

Abstract

The present application relates to the technical field of rail transit train operation control, and provides a group train tracking operation decision method and system considering communication delay. By constructing a lead vehicle-following vehicle distributed operation topology of the group train, each following vehicle receives the motion state information and prediction control sequence sent by the train in front through vehicle-to-vehicle communication; based on the distance, speed error between the following vehicle and the train in front, and the actual acceleration of the following vehicle, a discrete state space model is established, and a multi-objective optimization model is constructed in combination with constraints; the communication delay is discretized, and the delayed state of the train in front is compensated by using the prediction control sequence; the compensated state of the train in front and the current state of the following vehicle are fused, the optimal expected acceleration of the current control period is solved through the model, and a control instruction is generated. The above method and system can weaken the influence of communication delay on the acquisition of the state of the front vehicle, improve the accuracy of the tracking operation decision, and make the train tracking process more stable.

Description

Technical Field

[0001] This application relates to the field of rail transit train operation control technology, specifically to a group train tracking operation decision-making method and system that takes into account communication delay. Background Technology

[0002] As rail transit systems develop towards higher density and efficiency, train group operation based on virtual coupling is gradually becoming a feasible means to improve line throughput capacity. Under this operation mode, trains within the group obtain the operating status information of the preceding train through vehicle-to-vehicle communication and achieve train tracking control based on the preceding train's status.

[0003] Existing tracking control methods typically assume that the preceding train's status information can be transmitted to the following train in a timely and effective manner. However, in reality, factors such as track environment, signal obstruction, electromagnetic interference, and network transmission conditions can easily cause delays in train-to-train communication, resulting in a time lag in the following train's receipt of the preceding train's status information. Since the control system generates control commands based on this delayed status information, discrepancies can easily arise between the actual state of the preceding train and the state used by the following train, thus affecting the effectiveness of train tracking control.

[0004] Especially under close-interval tracking conditions, the lag in the status information of the preceding train will reduce the timeliness of the following train's response to changes in the preceding train's operation, thereby affecting the stability of train spacing control and operational safety. Summary of the Invention

[0005] This application aims to address the problem in the prior art where the lag in the status information of the preceding train reduces the accuracy of the following train's response to changes in the preceding train's operation. It provides a group train tracking operation decision-making method and system that takes into account communication delay, enabling the following train to respond accurately based on the current operating status of the preceding train.

[0006] To solve the above problems, this application is implemented as follows:

[0007] In a first aspect, this application provides a group train tracking operation decision-making method that takes into account communication delays, including: S1. Construct a distributed operation topology of lead car-follower car for the group of trains, so that each follower car can receive status data packets sent by the preceding train through car-to-car communication, wherein the status data packets include at least the motion status information and predictive control sequence of the preceding train; S2. Based on the spacing error and speed error between the following train and the preceding train, as well as the actual acceleration of the following train, a discrete state-space model of train tracking operation is established. Combined with train characteristic constraints, line speed limit constraints, and minimum safe tracking spacing constraints, a multi-objective optimization model is constructed with the goal of reducing spacing error and speed error and reducing control command fluctuations. S3. Discretize the communication delay corresponding to the state data packet, and compensate the lagging motion state of the preceding train due to the communication delay based on the predictive control sequence to obtain the compensated preceding train state. S4. The compensated preceding train state is fused with the real-time state of the following train to obtain the initial state vector of the current control cycle. Based on the multi-objective optimization model, the model predictive control rolling optimization solution is performed to output the optimal expected acceleration of the following train in the current control cycle. S5. Based on the optimal expected acceleration and combined with the train running resistance, a feedforward compensation calculation is performed to generate a longitudinal control force. According to the longitudinal control force, a control command is output to drive the following train to complete the tracking operation of the preceding train.

[0008] Secondly, this application provides a group train tracking operation decision-making system that takes into account communication delays, deployed in a group of trains, the group of trains including a lead car and at least one following car, the system comprising: The onboard status sensing unit is installed on the onboard end of each train to obtain the absolute position, actual running speed and actual acceleration of the train. The wireless communication device includes a vehicle-to-ground wireless communication terminal and a vehicle-to-vehicle direct communication terminal, wherein the vehicle-to-ground wireless communication terminal is used to receive global constraint information of group operation, and the vehicle-to-vehicle direct communication terminal is used to receive status data packets sent by the preceding train, the status data packets including at least the motion status information and predictive control sequence of the preceding train; An onboard safety computer, connected to the onboard state perception unit and the wireless communication device, is used to establish a discrete state-space model of train tracking operation based on the spacing error, speed error, and actual acceleration of the following vehicle and the preceding train. It also constructs a multi-objective optimization model with the goal of reducing spacing and speed errors and minimizing control command fluctuations, incorporating train characteristic constraints, line speed limit constraints, and minimum safe tracking spacing constraints. Furthermore, it discretizes the communication delay corresponding to the state data packet and performs integral compensation on the lagging motion state of the preceding train due to communication delay based on the predictive control sequence to obtain the compensated preceding train state. The compensated preceding train state is then fused with the real-time state of the following vehicle to obtain the initial state vector for the current control cycle. Finally, it performs rolling optimization of model predictive control based on the multi-objective optimization model to output the optimal expected acceleration of the following vehicle in the current control cycle. The control system, connected to the onboard safety computer, is used to perform feedforward compensation calculations based on the optimal expected acceleration and the train running resistance, generate longitudinal control force, and output control commands based on the longitudinal control force to drive the following vehicle to complete the tracking operation of the preceding train.

[0009] Compared with existing technologies that directly track and control based on the state of the preceding vehicle, this application has the following advantages: This application first discretizes the communication delay, then compensates for speed and position changes within the delay interval based on the predictive control sequence sent by the preceding train. This corrects the preceding train's position and speed, corresponding to historical moments, to a state closer to the current control cycle. Based on the gradual cumulative effect of acceleration on speed and speed on position during the train's longitudinal motion, the motion process of the preceding train within the delay period is discretized and reconstructed, rather than simply using outdated states or performing linear extrapolation. This reduces state mismatch caused by communication delay, improves the accuracy of the model's predictive control initial state, thereby reducing tracking errors and control fluctuations, and enhancing the effectiveness and engineering applicability of group train tracking operation decisions. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart provided in one embodiment of this application; Figure 2 This is an embodiment of the train group operation overall control model and hierarchical architecture diagram provided in this application; Figure 3 This is a schematic diagram of the most dangerous braking condition of adjacent trains within a group, provided in one embodiment of this application. Detailed Implementation

[0012] The technical solutions of this application 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 this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] The terms "first," "second," etc., used in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or apparatuses. Additionally, the use of "and / or" in this application indicates at least one of the connected objects, such as A and / or B and / or C, representing seven possibilities: including A alone, B alone, C alone, and the presence of both A and B, both B and C, both A and C, and the presence of A, B, and C.

[0014] In this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0015] The following describes the group train tracking operation decision-making method that takes into account communication delays provided in this application.

[0016] See Figure 1 , Figure 1 This is a flowchart illustrating a group train tracking operation decision-making method that takes into account communication delays, provided in an embodiment of this application. Figure 1 The group train tracking operation decision-making method shown can be executed by electronic devices, such as on-board equipment. This method addresses the collaborative operation requirements of group trains in virtual coupling scenarios. Under the condition of time delay disturbances in vehicle-to-vehicle communication, it achieves safe, stable, and high-precision tracking of the following train by constructing a distributed operation topology, establishing a predictive control model considering safety and stability, introducing a preceding train state delay compensation mechanism, and combining longitudinal control force feedforward calculation.

[0017] like Figure 1 As shown, the group train tracking operation decision-making method considering communication delays provided in this application may include the following steps: S1. Construct a distributed operation topology of lead car-follower car for the group of trains. Each follower car receives a status data packet sent by the preceding train through car-to-car communication. The status data packet includes at least the motion status information and predictive control sequence of the preceding train. Specifically, all trains in the group except the lead train operate as follower trains, each directly tracking its own preceding train. They receive status data packets sent by the preceding train through vehicle-to-vehicle communication. The motion status information in the status data packets may include the position and speed of the preceding train, and the predictive control sequence may include the predicted acceleration sequence corresponding to several future control steps.

[0018] S2. Based on the spacing error, speed error and actual acceleration of the following train and the preceding train, a discrete state-space model of train tracking operation is established. Combined with train characteristic constraints, line speed limit constraints and minimum safe tracking spacing constraints, a multi-objective optimization model is constructed with the goal of reducing spacing error and speed error and reducing control command fluctuations. In this step, spacing error, speed error, and actual acceleration of the following vehicle are selected as state variables. Considering the inertial response characteristics of the train traction and braking systems, the desired acceleration output by the controller cannot be instantaneously converted into the actual acceleration of the vehicle. Therefore, a state-space model with an execution inertial element is used to describe the longitudinal tracking process of the train, and a multi-objective optimization model is constructed.

[0019] S3. Discretize the communication delay corresponding to the state data packet, and perform integral compensation on the lagging motion state of the preceding train due to the communication delay based on the predictive control sequence to obtain the compensated preceding train state. In this step, there may be random or time-varying delays between the transmission of the status data packet from the preceding train to the receiving train. The information about the preceding train acquired by the following train in the current control cycle usually corresponds to an earlier time. If this delayed information is directly used for control solving, it can easily lead to a lag in the judgment of the preceding train's running trend, increasing the tracking error. Therefore, the communication delay is converted into discrete sampling steps, and the predictive control sequence in the status data packet is used to compensate for the speed and position changes of the preceding train during the delay period, thereby obtaining the preceding train's status corresponding to the current control cycle.

[0020] S4. The compensated state of the preceding train is fused with the real-time state of the following train to obtain the initial state vector of the current control cycle. Then, the rolling optimization solution of model predictive control is performed based on the multi-objective optimization model to output the optimal expected acceleration of the following train in the current control cycle. Specifically, after obtaining the initial state vector, based on the multi-objective optimization model constructed in step S2, model predictive control rolling optimization is performed within the set prediction and control time domains to obtain the optimal control sequence within several future control steps. In actual execution, only the first control variable in the optimal control sequence is output, and this is used as the optimal expected acceleration for the current control cycle. Upon entering the next sampling time, the optimization problem is reconstructed based on the latest state, and the above process is repeated.

[0021] S5. Based on the optimal expected acceleration and combined with the train running resistance, feedforward compensation calculation is performed to generate the target longitudinal control force. Control commands are output according to the target longitudinal control force to drive the following train to complete the tracking operation of the train in front.

[0022] In this step, the longitudinal control force is specifically the traction force or braking force, and the corresponding control is the traction control command or the braking control command.

[0023] This application addresses the control response deviations that exist in group train tracking operations under communication delay conditions by constructing a tracking operation decision-making method based on a distributed operating topology. This method utilizes state data packets to achieve local coordination at the adjacent train tracking level; at the control solution level, it integrates train characteristic constraints, line speed limit constraints, and minimum safe tracking distance constraints into the optimization process; at the delay compensation level, it uses predictive control sequences to correct the lagging state of the preceding train; and at the execution level, it incorporates running resistance for feedforward compensation calculations. Therefore, it can balance the compactness, stability, and safety of group train tracking operations in complex communication environments.

[0024] In one embodiment, the group train tracking operation decision-making method that takes into account communication delay is specifically applied to a group train operation control system, which includes at least a lead car, at least one following car, a car-to-car communication unit, an onboard safety computer, and a control system.

[0025] See Figure 2 , Figure 2 The overall control model and hierarchical architecture of train group operation are shown. Trains within the group are arranged according to the lead car, preceding car, etc. Following vehicle and following vehicle The sequence of trains forms a distributed operating topology, with adjacent trains exchanging status data packets via train-to-train communication. For example, the preceding train... To follow the car The train immediately preceding, following the train To follow the car The train immediately preceding it. The dashed box in the diagram indicates the train following it. The local tracking control unit includes a preceding train motion state compensation module, an error calculation and state construction module, a model predictive control module (MPC), and a lower-level execution control module. The error calculation and state construction module, represented by an addition / subtraction comparison node in the diagram, is used to compare the compensated preceding train state with that of the following train. The system compares the vehicle's status information with that of the following vehicle to calculate the spacing error and speed error, and then combines this information with the following vehicle's status information. The actual acceleration is used to construct the initial state quantities required for model predictive control. The Model Predictive Control (MPC) module performs rolling optimization based on the initial state quantities to obtain the optimal desired acceleration for the current control cycle. The lower-level execution control module generates corresponding traction control commands or braking control commands based on the optimal desired acceleration to drive the following vehicle. Complete the tracking operation of the train immediately preceding it.

[0026] The specific process for combining control with hardware is as follows: S1. Construct a distributed operation topology of the lead car-follower car group and a hierarchical collaborative control architecture. Each follower car receives a status data packet sent by the preceding train through the car-to-car communication unit. The status data packet includes the motion status information and predictive control sequence of the preceding train. The hierarchical collaborative control architecture includes an upper-level controller set in the on-board safety computer and a lower-level controller set in the control system. S2. The upper-level controller establishes a discrete state-space model of train tracking operation based on the spacing error, speed error and actual acceleration of the following train and the distance error between the following train and the preceding train. It also constructs a multi-objective optimization model by combining train characteristic constraints, line speed limit constraints and minimum safe tracking distance constraints. S3. The upper-level controller discretizes the communication delay corresponding to the status data packet and performs integral compensation on the lagging motion state of the preceding train due to the communication delay based on the predictive control sequence to obtain the compensated preceding train state corresponding to the current control cycle. S4. The upper-level controller fuses the compensated state of the preceding train with the real-time state of the following train to obtain the initial state vector of the current control cycle. Based on the multi-objective optimization model, it performs rolling optimization of model predictive control to solve the problem and outputs the optimal expected acceleration of the following train in the current control cycle. S5. The lower-level controller receives the optimal expected acceleration and performs feedforward compensation calculations in conjunction with the train running resistance to generate the target traction force or target braking force (i.e., the target longitudinal control force). The target traction force or target braking force is then converted into traction control commands or braking control commands to drive the following train to complete the tracking operation of the train in front.

[0027] In one embodiment, each following vehicle receives a status data packet sent by the preceding train based on vehicle-to-vehicle communication. The motion status information in the status data packet includes the position and speed information of the preceding train.

[0028] Specifically, the lead vehicle receives global constraint information related to group operation via vehicle-to-ground wireless communication, processes the global constraint information to form a corresponding global target state, and broadcasts it within the group. This part can adopt existing implementation methods. Each following vehicle receives status data packets sent by the preceding train via vehicle-to-vehicle communication. The motion state information in the status data packet includes at least the current position and current speed information of the preceding train.

[0029] In other words, the information sources in group operation control can be divided into two categories: global target information and local tracking information. The former is used to maintain the consistency of the group's operational goals, while the latter is used to ensure the real-time tracking control of adjacent trains. By dividing information, the direct dependence of each following train on the remote central node can be reduced, and deviations from the overall group goal can be avoided by operating solely based on local information.

[0030] In one embodiment, the communication delay discretization processing and immediate train motion state integral compensation in S3, the model predictive control rolling optimization solution in S4, and the longitudinal control force calculation in S5 are all performed within a fixed sampling period. and unified clock reference Run the following, among which, This represents the sampling sequence number corresponding to the discrete sampling time. For the corresponding discrete sampling time.

[0031] Specifically, to ensure that communication, computation, and execution operate within the same time frame, the entire control system employs a fixed sampling period. As the basic timing unit, and with As a unified clock reference, at each discrete sampling moment, the following train sequentially completes the following steps: receiving state data packets, discretizing communication delays, compensating for the motion state of the preceding train, constructing initial state vectors, solving model predictive control rolling optimization, and calculating longitudinal control force. With a unified clock reference, communication delays can be directly mapped to the number of discrete sampling steps, facilitating a unified description with both the prediction and control time domains. This also reduces state misalignment and control deviations caused by inconsistent time references between different functional modules.

[0032] In one embodiment, the process of establishing the discrete state-space model of train tracking operation in step S2 is as follows: From the perspective of actual operation control, the upper-level controller selects the distance error between the following vehicle and the vehicle in front. Speed ​​error and the actual acceleration of the following vehicle itself. As state variables, and forming a vector ,Right now:

[0033] and:

[0034]

[0035] in, and The sampling time of the preceding train respectively Position and velocity, and The following vehicle at the sampling time Position and velocity, For the length of the train, This represents the desired distance between the two trains.

[0036] Considering the first-order inertial response characteristics of the train traction and braking systems, when the upper-level controller issues the desired acceleration... At that time, the actual acceleration of the following vehicle satisfy:

[0037] in, To implement the inertial time constant.

[0038] Based on the above relationships, after discretizing the system, a discrete state-space model of train tracking operation can be obtained:

[0039] in, The system state matrix, To control the input matrix, The perturbation input matrix is... For the acceleration disturbance of the preceding train, each discrete matrix can be determined according to the sampling period. and execution inertial time constant It is obtained by discretizing a continuous model.

[0040] Based on this, a multi-objective optimization cost function is constructed with the objectives of reducing spacing and velocity errors and suppressing control input increment fluctuations. The cost function of the multi-objective optimization model is as follows. for:

[0041] in, This represents the sampling sequence number corresponding to the discrete sampling time. To predict the time domain, To control the time domain, For the prediction step number, This is the state variable vector corresponding to the prediction step. The state variable vector includes the spacing error, velocity error, and the actual acceleration of the following vehicle. To control the input increment, Here is the weight matrix for the state error penalty term. This is the weight matrix for the stationarity penalty term.

[0042] In one embodiment, during the rolling optimization process of model predictive control, at least train characteristic constraints, line speed limit constraints, and minimum safe following distance constraints are considered to ensure that the following train can still stably follow and meet operational safety requirements under communication delay conditions. The specific constraints are as follows.

[0043] The train characteristic constraints are:

[0044] The speed limit constraint for the line is:

[0045] The minimum safe tracking distance constraint is:

[0046] in, Assign a number to the following vehicle. This refers to the train number immediately preceding the following train. Indicates the first The optimal expected acceleration of the train following the vehicle in the current control cycle. For simplicity and to avoid confusion, it is denoted as... , To predict the expected acceleration of the vehicle during the predicted step, The minimum expected acceleration allowed for following vehicle. The maximum expected acceleration allowed for the following vehicle. To follow the vehicle at the predicted position , To predict the speed of the vehicle in the predicted step, To follow the vehicle at the predicted position of the predicted step, To determine the predicted position of the preceding train in the predicted step, For the length of the train, This is the minimum safe tracking distance.

[0047] Among these constraints, train characteristic constraints ensure that control inputs do not exceed the vehicle's traction and braking capabilities. Track speed limit constraints ensure that trains meet permissible speed requirements at different track locations. Minimum safe following distance constraints ensure that adjacent trains do not rear-end each other under normal operating conditions and sudden braking conditions.

[0048] Furthermore, in one embodiment, see [link to embodiment]. Figure 3 , Figure 3This diagram illustrates the speed-distance relationship between adjacent trains within a group under the most dangerous braking condition. This most dangerous braking condition can be defined as the preceding train being in its most advantageous braking condition and the following train in its most unfavorable braking condition. In the diagram, point H represents the position where the preceding train begins braking; H to P represent the most advantageous braking distance of the preceding train; and point P represents the stopping position of the preceding train. Point A represents the initial running position of the following train corresponding to the moment the preceding train begins braking; B to G represent the most unfavorable braking distance of the following train; and point G represents the stopping position of the following train. The distance between G and the rear end of the preceding train is a safety margin. A to B represent the delayed running stage from the moment the preceding train begins braking until the following train receives the deceleration information from the preceding train; the corresponding distance is... B to C represent the reaction phase of the following vehicle's onboard equipment; C to D represent the traction or brake cut-off phase; D to E represent the emergency braking idle phase; E to F represent the emergency braking establishment phase; and F to G represent the emergency braking implementation phase. If the following vehicle does not rear-end the preceding vehicle under this most dangerous braking condition, then the safe tracking requirements of adjacent trains within the group can be met under other conditions.

[0049] Therefore, minimum safe tracking distance Determined based on the most favorable braking distance of the preceding vehicle and the most unfavorable braking distance of the following vehicle, and satisfying the following conditions:

[0050]

[0051]

[0052] in, The braking distance of the following vehicle under the most unfavorable conditions. This is the braking distance of the vehicle in front under the most favorable conditions. To maintain a safe distance between trains, This refers to the distance traveled by the following vehicle between the time the preceding vehicle begins braking and the time the following vehicle begins braking. The initial running speed of the vehicle in front before braking; These represent the train travel distances of the following train in stages BC, CD, DE, EF, and FG, corresponding to the five stages with the most unfavorable braking conditions for the following train.

[0053]

[0054]

[0055]

[0056]

[0057] in, The initial running speed before the following vehicle brakes; For the response time of the onboard equipment; The time for the following vehicle to cut off traction (or braking); This is for the time it takes for the following vehicle to brake and travel unnecessarily. Establish time for the following vehicle to brake in an emergency; This is the maximum traction acceleration; For emergency braking deceleration; The value is a set value, preferably 0.5, which means that 50% of the train's emergency braking deceleration is used as the train's average deceleration during the EF phase.

[0058] This refers to the distance traveled by the following vehicle from the moment the preceding vehicle begins braking to the moment the following vehicle begins braking. To avoid confusion with the reaction time of the following vehicle's onboard equipment, the pre-processing time for information transmission during this phase is denoted as... The vehicle-to-vehicle communication delay time is recorded as ,but It can be represented as:

[0059] Compared to a single fixed-delay scenario or ideal communication conditions, this invention introduces random communication delays into the following train control framework to simulate a real train communication environment. Directly utilizing information from the preceding train obtained through vehicle-to-vehicle communication for control results in the following train always acquiring the historical state of the preceding train due to the objective existence of communication delays. Based on this lagging information, train operation control leads to increased group tracking spacing, restricting operational efficiency and even causing safety issues such as train collisions.

[0060] In one embodiment of the present invention, S3 specifically includes: The communication delay corresponding to the status data packet Discretize into time steps ,satisfy:

[0061] and, , ; The train ahead The predicted acceleration in the predictive control sequence obtained by solving at time step is denoted as . Multi-step integral compensation is performed on the position and speed of the preceding train that lags behind due to communication delays to obtain the compensated speed of the preceding train. and the position of the immediate preceding train after compensation ,satisfy:

[0062]

[0063] in, This is the sequence number of the summation step in the integral compensation process; when When greater than 0, and It can be obtained recursively from the transmission time state and predictive control sequence in the status data packet.

[0064] In addition, the status data packet carries a sending timestamp, which the following vehicle uses to calculate the communication delay when receiving the status data packet. .

[0065] The state of the preceding train after compensation is determined based on the compensated speed and position of the preceding train.

[0066] In one embodiment, S4 specifically includes: The current time is calculated based on the compensated status of the preceding train and the real-time status of the following train. The actual spacing error and speed error are combined with the actual acceleration of the following vehicle to form the initial state vector. ; In the set prediction time domain and control time domain Inside, with the initial state vector Using these as initial conditions, a finite-time-domain optimization solution model is constructed by combining a multi-objective optimization model with constraints on train characteristics, track speed limits, and minimum safe following distance. The finite-time domain optimization solution model is transformed into a quadratic programming problem and solved online to obtain the optimal control input increment sequence in the future control time domain. Extract the first control input increment from the optimal control input increment sequence and combine it with the control quantity from the previous control cycle to generate the optimal expected acceleration for the current control cycle.

[0067] Specifically, step 4.1, state initialization: The compensated preceding train state obtained in S3 is fused with the real-time state of the current train obtained by the onboard state perception unit to calculate the current time. The actual distance error and speed error are combined with the actual acceleration of the following vehicle to form the initial state vector of the prediction system. ; Step 4.2: Establish the solution model: In the defined prediction time domain and control time domain Internally, starting from the initial state vector, and combining it with the set multi-objective optimization cost function, A finite-time-domain optimization model for train tracking operation is constructed, incorporating constraints such as train characteristic constraints, track speed limit constraints, and minimum safe following distance constraints. The optimization variable is the control input increment sequence, which satisfies the following conditions:

[0068] in, .

[0069] Step 4.3, Quadratic Programming Solution: In each discrete control cycle Within this framework, the finite-time optimization model is transformed into a standard quadratic programming problem for online solution, yielding the optimal control input increment sequence within the future control time domain:

[0070] Step 4.4, Rolling Time Domain Execution: Extract only the first control input increment from the optimal control input increment sequence. And combined with the control quantity of the previous control cycle Generate the optimal desired acceleration for the current control cycle. ,Right now:

[0071] The optimal desired acceleration is sent to the lower-level controller for execution. When entering the next discrete control cycle... At that time, the states of the preceding vehicle and the current vehicle are reacquired, and the prediction time domain is rolled forward by one sampling period, and the above process is repeated.

[0072] In one embodiment, the lower-level controller receives the optimal desired acceleration and actively calculates the target traction force or target braking force (i.e., longitudinal control force) through a feedforward compensation strategy to counteract the nonlinear resistance during train operation and drive the train to perform physical tracking operation.

[0073] The train's running resistance exhibits highly nonlinear and time-varying characteristics. Directly inputting it into the prediction domain of the upper-level MPC would result in a huge computational load and cause severe computational delays. To address this, a feedforward compensation strategy is introduced in the lower-level TCMS controller. The lower-level controller receives the desired acceleration. The system uses onboard sensors to acquire the train's real-time speed and the gradient and curvature of the current track, and calculates the target traction or braking force to be applied. The feedforward compensation control law is as follows:

[0074] in, For train quality, The optimal expected acceleration for the current control cycle, As the basic resistance, Add resistance to the ramp, Adding resistance to the curve. Through this feedforward control law, the train's complex nonlinear running resistance is offset. Ultimately, TCMS will... It is converted into a level control signal for the traction converter or braking system to drive the train to perform physical tracking.

[0075] One embodiment of this application provides a group train tracking operation decision-making system that takes into account communication delays. It includes: The onboard status sensing unit is installed on the onboard end of each train to obtain the absolute position, actual running speed and actual acceleration of the train. The wireless communication equipment includes a vehicle-to-ground wireless communication terminal and a vehicle-to-vehicle direct communication terminal. The vehicle-to-ground wireless communication terminal is used to receive global constraint information of group operation, and the vehicle-to-vehicle direct communication terminal is used to receive status data packets sent by the preceding train. The status data packets include at least the motion status information and predictive control sequence of the preceding train. An onboard safety computer, connected to an onboard state perception unit and wireless communication equipment, is used to establish a discrete state-space model of train tracking operation based on the spacing error, speed error, and actual acceleration of the following vehicle and the preceding train. It also constructs a multi-objective optimization model, combining train characteristic constraints, line speed limit constraints, and minimum safe tracking spacing constraints, with the goal of reducing spacing and speed errors and minimizing control command fluctuations. Furthermore, it discretizes the communication delay corresponding to the state data packets and performs integral compensation on the lagging motion state of the preceding train due to communication delay based on the predictive control sequence, obtaining the compensated preceding train state. The compensated preceding train state is then fused with the real-time state of the following vehicle to obtain the initial state vector for the current control cycle. Finally, it performs rolling optimization of model predictive control based on the multi-objective optimization model, outputting the optimal expected acceleration of the following vehicle in the current control cycle. The control system, connected to the onboard safety computer, is used to perform feedforward compensation calculations based on the optimal expected acceleration and the train running resistance, generate longitudinal control force, and output control commands based on the longitudinal control force to drive the following vehicle to complete the tracking operation of the preceding train.

[0076] The system can achieve the functions described in this application. Figure 1 The various processes in the method embodiments, and the ways to achieve the same beneficial effects, will not be repeated here to avoid repetition.

[0077] This application also provides an electronic device, including a processor and a memory, wherein the memory stores a program or instructions that can run on the processor. When the program or instructions are executed by the processor, they implement the various steps of the above-described embodiment of the group train tracking operation decision method considering communication delay, and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0078] It should be noted that the electronic device in this application may be an in-vehicle device.

[0079] This application also provides a readable storage medium storing a program or instructions that, when executed by a processor, implement the various processes of the above-described embodiment of the group train tracking operation decision method considering communication delay, and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0080] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (Read-Only Memory). Only memory (ROM), random access memory (RAM), magnetic disks or optical disks, etc.

[0081] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0082] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A group train tracking operation decision-making method considering communication delay, characterized in that, include: S1. Construct a distributed operation topology of lead car-follower car for the group of trains, so that each follower car receives a status data packet sent by the preceding train, wherein the status data packet includes at least the motion status information and predictive control sequence of the preceding train; S2. Based on the spacing error and speed error between the following train and the preceding train, as well as the actual acceleration of the following train, a discrete state-space model of train tracking operation is established. Combined with train characteristic constraints, line speed limit constraints, and minimum safe tracking spacing constraints, a multi-objective optimization model is constructed with the goal of reducing spacing error and speed error and reducing control command fluctuations. S3. Discretize the communication delay corresponding to the state data packet, and compensate the lagging motion state of the preceding train due to the communication delay based on the predictive control sequence to obtain the compensated preceding train state. S4. The compensated preceding train state is fused with the real-time state of the following train to obtain the initial state vector of the current control cycle. Based on the multi-objective optimization model, the model predictive control rolling optimization solution is performed to output the optimal expected acceleration of the following train in the current control cycle. S5. Based on the optimal expected acceleration and combined with the train running resistance, a feedforward compensation calculation is performed to generate a longitudinal control force. According to the longitudinal control force, a control command is output to drive the following train to complete the tracking operation of the preceding train.

2. The group train tracking operation decision-making method considering communication delay according to claim 1, characterized in that, The discretization process and integral compensation of the preceding train's motion state in S3, the rolling optimization solution of model predictive control in S4, and the generation of longitudinal control force in S5 are all performed within a fixed sampling period. and unified clock reference Run it below, where, This represents the sampling sequence number corresponding to the discrete sampling time. For discrete sampling times.

3. The group train tracking operation decision-making method considering communication delay according to claim 1, characterized in that, In S2, the cost function of the multi-objective optimization model for: in, This represents the sampling sequence number corresponding to the discrete sampling time. To predict the time domain, To control the time domain, For the prediction step number, The state variable vector corresponding to the prediction step includes at least the spacing error, velocity error, and the actual acceleration of the following vehicle. To control the input increment, The weight matrix is ​​the weight matrix for the state error penalty term. This is the weight matrix for the stationarity penalty term.

4. The group train tracking operation decision-making method considering communication delay according to claim 2, characterized in that, In S2, the train characteristic constraint is expressed as: The speed limit constraint for the line is expressed as follows: The minimum safe tracking distance constraint is expressed as: in, Assign vehicle number to the following vehicle. This refers to the train number immediately preceding the following train. To predict the expected acceleration of the vehicle during the predicted step, The minimum expected acceleration allowed for following vehicle. The maximum expected acceleration allowed for the following vehicle. To follow the vehicle at the predicted position , To predict the speed of the vehicle in the predicted step, To follow the vehicle at the predicted position of the predicted step, To determine the predicted position of the preceding train in the predicted step, For the length of the train, This is the minimum safe tracking distance.

5. The group train tracking operation decision-making method considering communication delay according to claim 4, characterized in that, In S2, the minimum safe tracking distance Determined based on the most favorable braking conditions of the preceding vehicle and the most unfavorable braking conditions of the following vehicle, and satisfying the following: in, This is the braking distance of the following vehicle under the most unfavorable conditions. This is the braking distance of the vehicle in front under the most favorable conditions. To maintain a safe distance between trains, This refers to the distance traveled by the following vehicle between the time the preceding vehicle begins braking and the time the following vehicle begins braking.

6. The group train tracking operation decision-making method considering communication delay according to claim 2, characterized in that, S3 specifically includes: The communication delay corresponding to the status data packet Discretize into time steps ,satisfy: The train ahead The predicted acceleration in the predictive control sequence obtained by solving at time step is denoted as . Multi-step integral compensation is performed on the position and speed of the preceding train that lags behind due to communication delays to obtain the compensated speed of the preceding train. and the position of the immediate preceding train after compensation ,satisfy: in, This refers to the sequence number of the summation step in the integral compensation process; This refers to the prediction step number; The state of the preceding train after compensation is determined based on the compensated speed and position of the preceding train.

7. The group train tracking operation decision-making method considering communication delay according to claim 1, characterized in that, S4 specifically includes: The current time is calculated based on the compensated status of the preceding train and the real-time status of the following train. The actual spacing error and speed error are combined with the actual acceleration of the following vehicle to form the initial state vector. ; In the set prediction time domain and control time domain Within, with the initial state vector Starting from the above, a finite time domain optimization solution model is constructed by combining the multi-objective optimization model with the train characteristic constraints, track speed limit constraints and minimum safe following distance constraints. The finite-time-domain optimization solution model is transformed into a quadratic programming problem and solved online to obtain the optimal control input increment sequence in the future control time domain. Extract the first control input increment from the optimal control input increment sequence and combine it with the control quantity from the previous control cycle to generate the optimal expected acceleration for the current control cycle.

8. The group train tracking operation decision-making method considering communication delay according to claim 1, characterized in that, In S5, the longitudinal control force satisfy: in, For train quality, The optimal expected acceleration for the current control cycle. As the basic resistance, Add resistance to the ramp, Add resistance to the curve.

9. A group train tracking operation decision-making system considering communication delay, characterized in that, Deployed in a group of trains, the group of trains including a lead car and at least one following car, the system includes: The onboard status sensing unit is installed on the onboard end of each train in the group of trains to obtain the absolute position, actual running speed and actual acceleration of the train. A vehicle-to-vehicle direct communication terminal is used to receive status data packets sent by the preceding train. The status data packets include at least the motion status information and predictive control sequence of the preceding train. The onboard safety computer, connected to the onboard state perception unit and the vehicle-to-vehicle direct communication terminal, is used to establish a discrete state-space model of train tracking operation based on the spacing error, speed error, and actual acceleration of the following vehicle and the preceding train. It also constructs a multi-objective optimization model with the goal of reducing spacing and speed errors and minimizing control command fluctuations, incorporating train characteristic constraints, line speed limit constraints, and minimum safe tracking spacing constraints. Furthermore, it discretizes the communication delay corresponding to the state data packet and performs integral compensation on the lagging motion state of the preceding train due to communication delay based on the predictive control sequence to obtain the compensated preceding train state. The compensated preceding train state is then fused with the real-time state of the following vehicle to obtain the initial state vector for the current control cycle. Finally, it performs rolling optimization of model predictive control based on the multi-objective optimization model to output the optimal expected acceleration of the following vehicle in the current control cycle. The control system, connected to the onboard safety computer, is used to perform feedforward compensation calculations based on the optimal expected acceleration and the train running resistance, generate longitudinal control force, and output control commands based on the longitudinal control force to drive the following vehicle to complete the tracking operation of the preceding train.

Images