A High-Speed ​​Train Adhesion Constraint Balance Control Method Based on Virtual Trajectory Projection

By using a control method based on virtual trajectory projection and auxiliary dynamic compensation, the problems of ideal trajectory exceeding the range and input saturation in the adhesion control of high-speed trains were solved. This method achieves constraint satisfaction and bounded trajectory tracking error under the condition of limited actuator input, thereby improving the safety and stability of train operation.

CN123104581APending Publication Date: 2026-11-17CHANGCHUN UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611569664.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-12
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing high-speed train adhesion control methods, the ideal virtual control trajectory may exceed the allowable range of the state, and input saturation affects control performance, making it difficult to effectively maintain wheel-rail anti-slip constraints and resulting in unbounded trajectory tracking errors.

Method used

A control structure combining virtual trajectory projection and auxiliary dynamic compensation is adopted. By constructing a smooth projection mapping, the ideal virtual control trajectory is projected to the allowable range. Command filtering and robust backstepping control are used to suppress the effects of projection deviation and input saturation deviation. A robust backstepping control law is constructed to generate traction or braking torque commands.

Benefits of technology

This technology enables the train to maintain its state while adhering to constraints and keeping the trajectory tracking error bounded, even when actuator input is limited. It avoids the impact of virtual trajectory exceeding the allowable range and input saturation on control performance, thereby improving the safety and stability of high-speed trains.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
  • Figure FT_2
    Figure FT_2
  • Figure FT_3
    Figure FT_3
Patent Text Reader

Abstract

This invention discloses a high-speed train adhesion constraint balance control method based on virtual trajectory projection, belonging to the field of high-speed train wheel-rail adhesion control technology, and is used to solve the conflict between anti-slip safety constraints and actuator input limitations. The method establishes a wheel-rail adhesion dynamics model with input saturation, setting displacement, velocity, and slip velocity constraints. For cases where the ideal virtual control trajectory generated in recursive control may exceed the allowable range of the corresponding state, a smooth, monotonic, and reversible projection mapping is constructed to map it to the allowable range with a reserved safety margin. An auxiliary dynamic system is used to recursively compensate for projection deviations and input saturation deviations, and traction or braking torque commands are generated by combining command filtering and robust backstepping control. This invention is used for adhesion control during the traction and braking processes of high-speed trains, ensuring that the train state meets preset constraints and that the trajectory tracking error remains bounded.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of high-speed train wheel-rail adhesion control technology, specifically relating to a high-speed train adhesion constraint balance control method based on virtual trajectory projection. Background Technology

[0002] The traction and braking of high-speed trains rely on the adhesion between the wheels and rails. Good wheel-rail adhesion characteristics are a crucial foundation for the effective transmission of traction and braking forces, ensuring the safe and stable operation of the train. Wheel-rail adhesion exhibits nonlinear characteristics, which are affected by factors such as temperature, humidity, wheel-rail surface roughness, and cleanliness. When wheel-rail adhesion conditions change, or when the traction and braking actions are incompatible with the current adhesion conditions, the wheelset may experience traction slippage or braking skidding, affecting the effective transmission of traction and braking forces and potentially exacerbating damage to the wheel-rail surface, thus impacting the safety and stability of train operation.

[0003] Existing adhesion control methods for high-speed trains include re-adhesion control and active anti-slip control. Re-adhesion control typically detects wheel slippage or coasting based on wheel speed difference, wheelset acceleration, or slip threshold, and returns the wheel-rail operating point to a stable adhesion region by reducing, maintaining, and gradually restoring the traction or braking torque amplitude. This type of method mainly relies on slip detection results for torque correction, and its intervention timing is related to the detection threshold and response process. Active anti-slip control reduces the risk of abnormal slippage by adjusting the wheel-rail slip state. One type of method identifies and tracks the optimal slip state corresponding to the adhesion peak to improve adhesion utilization; another type incorporates the safe operating range of slip ratio or slip speed as an explicit constraint into the controller design to keep the wheel-rail slip state within the allowable range.

[0004] Patent CN109484384A discloses an optimal adhesion braking control method and system for high-speed trains. It calculates the optimal creep rate by estimating track surface parameters and combines a sliding mode observer and a creep rate tracking controller to achieve estimation and tracking of the optimal creep rate. This scheme uses optimal creep rate tracking as the main control objective to maximize braking capability under different track surface conditions.

[0005] Patent CN110095979B discloses a high-speed train adhesion anti-slip control method based on an asymmetric barrier Lyapunov function. It employs a hybrid control structure combining an asymmetric barrier Lyapunov function controller and a sliding mode controller, and combines sliding mode observation and variable step size search to determine the desired creep speed, achieving creep speed tracking and anti-slip control. This scheme also considers re-adhesion control when the initial creep state is outside the feasible adhesion region.

[0006] Furthermore, for constraint control methods where some control laws explicitly depend on constraint boundaries, the required control action may increase as the slip state approaches the safety boundary, and some control structures may also exhibit boundary singularities. The actual traction or braking torque is limited by the actuator's output capability; when the torque command exceeds the allowable range, input saturation will cause inconsistencies between the actual torque and the command, thus affecting anti-slip constraint maintenance and trajectory tracking performance.

[0007] Therefore, it is urgent to establish a high-speed train adhesion constraint balance control method based on virtual trajectory projection, which can adjust the ideal virtual control trajectory that may exceed the allowable range of the corresponding state during the backstep recursion process, reduce the impact of input saturation on closed-loop control, and keep the train trajectory tracking error bounded while satisfying wheel-rail anti-slip constraints. Summary of the Invention

[0008] This invention addresses the problems in high-speed train adhesion control where the ideal virtual control trajectory may exceed the allowable range of the corresponding state, and input saturation affects control performance. It proposes a high-speed train adhesion constraint balance control method based on virtual trajectory projection. This method projects the ideal virtual control trajectory onto the corresponding allowable interval and compensates for the effects of projection deviation and input saturation deviation to satisfy train state constraints and maintain bounded trajectory tracking error.

[0009] This invention employs a control structure combining virtual trajectory projection and auxiliary dynamic compensation. In one specific embodiment, it is implemented by combining command filtering and robust backstepping control, following these steps:

[0010] Controller construction:

[0011] Step 1: Establish a wheel-rail adhesion dynamics model that couples the longitudinal motion of the high-speed train with the rotation of the driving wheelset. Use the train displacement, running speed and wheelset angular velocity as state variables, and traction or braking torque command as control input. Establish an actuator input saturation model.

[0012] Step 2: Based on the train operating section, allowable speed range, and wheel-rail anti-slip requirements, set displacement constraints, speed constraints, and slip speed constraints respectively, and convert the wheel-rail adhesion dynamics model into a third-order strict feedback form suitable for recursive control design;

[0013] Step 3: Construct a smooth, monotonic and reversible projection mapping based on the allowable range of each state, and project the ideal virtual control trajectory generated in the backstep recursion process onto the allowable range with a safety margin to obtain a feasible virtual control trajectory.

[0014] Step 4: Calculate the projection deviation between the ideal virtual control trajectory and the projected virtual control trajectory, and calculate the input saturation deviation based on the difference between the torque command and the actual output of the actuator; construct the auxiliary dynamic system using the projection deviation and the input saturation deviation as driving signals.

[0015] Step 5: Filter the projected virtual control trajectory using a command filter to obtain the corresponding filtered signal and its derivative, and construct the recursive tracking error by combining it with the auxiliary dynamic system state;

[0016] Step 6: Based on the recursive tracking error, the filtered signal, and the auxiliary dynamic system state, a robust backstepping method is used to construct the ideal virtual control trajectory and the traction or braking torque command in sequence, and to suppress the influence of system disturbances and model uncertainties on control performance.

[0017] Step 7: Obtain the high-speed train's operating status online, update the virtual trajectory projection, command filtering, auxiliary dynamic system, and torque command in sequence, and apply the actual torque after actuator saturation limiting to the drive wheelset.

[0018] The beneficial effects of this invention include: First, by projecting the virtual trajectory, the ideal virtual control trajectory is mapped to the corresponding allowable range with a reserved safety margin, eliminating the need to pre-guarantee that the ideal virtual control trajectory meets state constraints and avoiding the requirement for the constrained state to track a virtual reference that exceeds the allowable range; Second, by using auxiliary dynamic compensation, the influence of virtual trajectory projection deviation and input saturation deviation on each stage of the control process is reduced; Third, in the implementation using command filtering and robust backstepping control, repeated analytical differentiation of the virtual control signal is avoided during the backstepping recursion process, and the influence of model uncertainty and external disturbances on control performance is suppressed; Fourth, when the corresponding initial conditions and control design conditions are met, the train state can meet preset constraints under the condition of limited actuator input, and the trajectory tracking error is kept bounded, providing technical support for wheel-rail adhesion control during the traction and braking process of high-speed trains. Attached Figure Description

[0019] Figure 1 This is a general framework diagram of the high-speed train adhesion constraint balance control method based on virtual trajectory projection described in this invention;

[0020] Figure 2 This is a graph showing the trajectory tracking results of high-speed train displacement and speed.

[0021] Figure 3 The graph shows the changes in slip velocity and slip ratio under anti-slip constraints.

[0022] Figure 4 A graph showing the variation between the torque command and the actual output torque under actuator input saturation conditions;

[0023] Figure 5 This is a graph showing the changes in virtual trajectory projection deviation and input saturation deviation.

[0024] Figure 6 The result diagram is used to track errors and assist in the dynamic changes of the system state. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0026] This embodiment provides a high-speed train adhesion constraint balance control method based on virtual trajectory projection. For example... Figure 1 The diagram shown is an overall framework diagram of the method of the present invention; Figure 2 This is a graph showing the results of train displacement and speed tracking. Figure 3 The graph shows the results of the wheelset slip speed and slip ratio changes; Figure 4 This is a graph showing the torque command and the actual torque result after input saturation; Figure 5 The graph shows the results of the virtual trajectory projection deviation and input saturation deviation changes. Figure 6 To track errors and provide a graph illustrating the dynamic changes in the system state, the specific implementation steps are as follows, based on the above framework:

[0027] Step 1: Establish a longitudinal dynamics model of a high-speed train that includes wheel-rail adhesion nonlinearity, running resistance, external disturbances, and actuator input saturation.

[0028] The longitudinal motion and rotational dynamics of the drive wheelset of a high-speed train are expressed as follows:

[0029] (1)

[0030] in, , and These represent the train's displacement, speed, and wheel-to-wheel angular velocity, respectively. Indicates train mass; and These represent the wheelset's moment of inertia and viscous damping coefficient, respectively. This indicates the traction or braking torque command generated by the controller. This indicates the actual output torque of the actuator; , and This represents external disturbances and unmodeled dynamics.

[0031] Wheel-rail slip ratio is defined as:

[0032] (2)

[0033] in, Indicates the wheel-rail slip speed. Indicates the creep reference velocity. Indicates the wheel radius. When... When corresponding to traction conditions, when The corresponding braking condition.

[0034] Wheel-rail adhesion coefficient, adhesion force, adhesion torque, and train running resistance are expressed as follows:

[0035] (3)

[0036] in, , and This represents the adhesion characteristic parameters related to wheel-rail contact conditions. , and This represents the unknown train running resistance parameter. Considering the torque output capacity of the traction and braking actuators, the actual output torque is expressed as:

[0037] (4)

[0038] in, and These represent the upper and lower limits of the actuator's allowed output torque, respectively.

[0039] Step 2: Determine the constraints on train operation status and wheel-rail anti-slip constraints, and transform the train dynamics model into a third-order strict feedback form.

[0040] Train displacement, running speed, and wheel-rail slip speed satisfy the following:

[0041] (5)

[0042] in, . This indicates the preset upper limit of the safe slip ratio. This represents a small positive number used to avoid sensitivity calculations for glide rate under low-speed conditions. Let... , , Equation (1) can be transformed into:

[0043] (6)

[0044] in, , ; This represents a nonlinear function determined by the train dynamics model; This represents the lumped uncertainty term formed by external disturbances, parameter uncertainties, and adhesion model errors; This indicates the corresponding control gain.

[0045] To give one feasible way of taking each function in equation (6), let and order ,by This represents the adhesion coefficient function calculated based on the wheel-rail adhesion characteristic parameters set in the controller. According to the mean-value theorem, within the safe creep region, the actual adhesion effect can be equivalently expressed as... And select the relevant running speed Smooth positive function As the control gain of the second state equation in equation (6), the nonlinear function and the control gain can be taken as follows:

[0046] (7)

[0047] The corresponding lumped uncertainty term can be taken as:

[0048] (8)

[0049] in, , , and These represent the adhesion coefficient function set by the controller, the actual adhesion coefficient function, the equivalent adhesion gain within the safe creep region, and the smooth positive control gain, respectively. In equation (8), the unknown parameters... , , The resulting operating resistance effects are respectively included in the lumped uncertainty terms. and In this model, adhesion model errors and external disturbances are also included in the corresponding lumped uncertainty terms. The controller performs robust compensation based on a conservative upper bound of the lumped uncertainty terms, without needing to obtain... The exact value. Assuming... Within the preset state's allowable range, the control gain is continuously bounded and... If the expression is positive and has known upper and lower bounds, then:

[0050] (9)

[0051] To facilitate unified description and control design, the control design range for the three states is represented as follows:

[0052] (10)

[0053] in, , and These represent train displacement, running speed, and wheel-rail slip speed, respectively.

[0054] In this embodiment, the initial position of the train is taken as the origin of the displacement coordinate system. To ensure that the initial displacement lies within the control analysis interval, an auxiliary lower boundary is introduced. ,in Appropriate reference trajectory margins and control parameters are selected to ensure that the distance between the reference displacement and the analysis boundary can cover the influence of displacement tracking error. This auxiliary lower boundary is used for control analysis, does not change the actual initial position of the train, and is not considered as the actual allowable backtracking distance.

[0055] The remaining boundaries are determined according to equation (5). The parameters of each boundary are:

[0056] (11)

[0057] Step 3: Construct a virtual trajectory projection mapping to map the ideal virtual control trajectory generated during the backstep recursion process to the allowable range of the corresponding state.

[0058] For the ideal virtual control trajectory , Construct projection mapping To satisfy:

[0059] (12)

[0060] in, This represents the reserved safety margin. The projection mapping adopts:

[0061] (13)

[0062] in, . This represents the projection shape parameter. The projection mapping is smooth, strictly monotonic, and invertible, and its first derivative is bounded. It can map the ideal virtual control trajectory to an allowable range after reserving a safety margin without requiring the ideal virtual control trajectory to meet state constraints in advance.

[0063] Step 4: Calculate the virtual trajectory projection deviation and input saturation deviation, and construct an auxiliary dynamic system to compensate for the impact of the two types of deviations in the recursive control process.

[0064] Virtual trajectory projection deviation and input saturation deviation are defined as follows:

[0065] (14)

[0066] The above deviations are combined into a vector. And construct an auxiliary dynamic system:

[0067] (15)

[0068] in, To assist in the dynamic system state, , This represents the attenuation gain of the auxiliary dynamic system. This represents the smoothing parameter of the arctangent compensation term and specifies the terminal boundary. In equation (15), , and Enter in sequence , and The dynamic equations; simultaneously, the subsequent auxiliary state Entering the previous level of auxiliary dynamics, the virtual trajectory projection deviation and input saturation deviation are transmitted along the strict feedback structure and recursively compensated.

[0069] Step 5: Set the command filter and construct the recursive tracking error based on the command filter output and the auxiliary dynamic system state.

[0070] Set a first-order command filter for the projected virtual control trajectory:

[0071] (16)

[0072] in, Indicates the output of the command filter. This represents the command filter time constant, and is related to the time constant in equation (15). The meanings are different. The corresponding command filtering error is defined as:

[0073] (17)

[0074] Based on the auxiliary dynamic system state, construct the recursive tracking error:

[0075] (18)

[0076] in, This represents the desired displacement trajectory of the train. The command filter is used to obtain the projected virtual trajectory and its dynamic information, avoiding repeated analytical differentiation during the backstep recursion process.

[0077] Step 6: Construct a robust backstepping control law based on the recursive tracking error, and sequentially generate the first virtual control trajectory, the second virtual control trajectory, and the traction or braking torque command.

[0078] Define intermediate functions:

[0079] (19)

[0080] First Virtual Control Trajectory :

[0081] (20)

[0082] Second Virtual Control Trajectory :

[0083] (twenty one)

[0084] Traction or braking torque command Represented as:

[0085] (twenty two)

[0086] in, Indicates the error feedback gain. The robust smoothing parameter is represented. The arctangent terms in equations (20) to (22) are used to suppress the influence of lumped uncertainty and adjacent recursive error terms on control performance.

[0087] Step 7: Set controller parameters and safety margins to ensure that the closed-loop signal is bounded and that the train status meets the corresponding constraints.

[0088] Choose a composite Lyapunov function:

[0089] (twenty three)

[0090] For the control law used in this embodiment, under the conditions described in equation (9), it is assumed that the virtual trajectory projection deviation and the input saturation deviation satisfy the upper bound of the given amplitude. A positive feedback gain and smoothing parameter are selected, and the command filter time constant and the auxiliary dynamic system attenuation gain are set to satisfy the following parameter conditions:

[0091] (twenty four)

[0092] in, , These are the lower and upper bounds of the control gain, respectively. The time constant for the command filter. To assist in attenuating the gain of the dynamic system; , The attenuation margin design constant is positive. Under the condition that the initial state is within the allowable range, the desired trajectory is within the displacement constraint, and the deviation... , and When the parameters are bounded and the filter parameters, auxiliary dynamic system parameters, and feedback gain satisfy the stability condition, we have:

[0093] (25)

[0094] in, , , and These are constants determined by the controller parameters and the upper bound of the deviation. These constants are determined by Lyapunov estimation of the control parameters, the upper and lower bounds of the control gain, the upper bound of the deviation, and the upper bound of the time derivative of the projected trajectory; they are not arbitrarily selected as independent design parameters. Simultaneously, a projection safety margin is selected. satisfy:

[0095] (26)

[0096] in, , and These represent the upper bounds of the command filtering error, recursive tracking error, and auxiliary dynamic system state, respectively. In addition to being located within the corresponding control analysis interval, the initial state also satisfies... , The aforementioned upper bound is determined under the initial conditions according to equation (25) and the command filter error equation, and includes the influence of the initial transient.

[0097] Online closed-loop control

[0098] Step 1: Collect the train's current displacement, running speed, and wheelset angular velocity, and calculate the slip speed, slip ratio, and the state constraint boundary at the current moment.

[0099] Data collected during each control cycle , and Calculate according to formula (2) and The slip velocity safety boundary is updated according to equations (5) and (11). The command filter state of the previous control cycle is read. , and auxiliary dynamic system status , , .

[0100] Step 2: Calculate the two-level ideal virtual control trajectory according to the strict feedback recursion order, and project the virtual trajectory respectively.

[0101] First, calculate according to equation (18). And obtain the ideal virtual control trajectory according to equations (19) and (20). The projection trajectory is then obtained using equation (13). And calculate the projection deviation according to equation (14). Further calculations The ideal virtual control trajectory is obtained according to equations (19) and (21). The projection trajectory is obtained through equation (13). And calculate the projection deviation. .

[0102] Step 3: Calculate the traction or braking torque command and generate the input saturation deviation.

[0103] Calculated based on the current command filter output and the auxiliary dynamic system state. Then, the torque command is generated according to equations (19) and (22). By inputting the torque command into the saturation circuit shown in equation (4), the actual output torque of the actuator can be obtained. Calculate the input saturation deviation according to formula (14). The actual output torque It is applied to the train's traction or braking actuator.

[0104] Step 4: Synchronously update the command filter and auxiliary dynamic system, and enter the next control cycle.

[0105] Projection trajectory and Input the command filters shown in equation (16) respectively; , and Together with the auxiliary dynamic system shown in the corresponding recursive tracking error input formula (15), synchronously update , and , , Therefore, in each control cycle, the calculation and projection of the first virtual control trajectory, the calculation and projection of the second virtual control trajectory, the generation of torque commands and the input saturation processing are completed sequentially, and the auxiliary dynamic system is used to recursively compensate for the projection deviation of the two-stage virtual trajectory and the input saturation deviation of the final stage.

[0106] The above embodiments of the present invention are merely illustrative of the technical solutions of the present invention and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is impossible to exhaustively list all possible implementations here. Any obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. A high-speed train adhesion constraint balance control method based on virtual trajectory projection, characterized in that, The method comprises two parts: controller construction and online closed-loop control. Controller construction: Step 1: Establish a wheel-rail adhesion dynamics model that couples the longitudinal motion of the high-speed train with the rotation of the drive wheelset, using traction or braking torque commands as control inputs, and establish an input saturation model based on the actuator output capacity; Step 2: Based on the train operating section, allowable speed range, and wheel-rail anti-slip requirements, set state constraints for displacement, running speed, and slip speed, and convert the wheel-rail adhesion dynamics model into a strict feedback form suitable for backstepping control design; Step 3: Construct a projection mapping based on the allowable range of running speed and sliding speed, and map the two-level ideal virtual control trajectories generated in the backstep recursion process to the corresponding allowable intervals with reserved safety margins to obtain feasible virtual control trajectories; Step 4: Calculate the projection deviation based on the difference between the ideal virtual control trajectory and the corresponding projected trajectory, and calculate the input saturation deviation based on the difference between the torque command and the actual output torque of the actuator; drive the corresponding auxiliary dynamic system with the two-level projection deviation and input saturation deviation, and make the subsequent auxiliary state enter the previous auxiliary dynamic state, and recursively compensate for the influence of the deviation. Step 5: Filter the feasible virtual control trajectory using a command filter to obtain the filtered signal and its derivative, and construct a recursive tracking error by combining the train status, the desired displacement trajectory, and the auxiliary dynamic system status. Step 6: Based on the recursive tracking error, the filtered signal and its derivative, and the auxiliary dynamic system state, robust backstepping control is used to construct two levels of ideal virtual control trajectory and traction or braking torque command in sequence; Step 7: Based on the initial state, desired displacement trajectory, and upper bounds of the amplitudes of command filtering error, recursive tracking error, and auxiliary dynamic system state, set the controller parameters and projection safety margin; Online closed-loop control: Step 1: Collect the train's current displacement, running speed, and wheelset angular velocity; calculate the slip speed, slip ratio, and current state constraint boundaries. Step 2: Calculate the two-level ideal virtual control trajectory in reverse recursive order, project the virtual trajectory respectively, and calculate the corresponding projection deviation; Step 3: Calculate the traction or braking torque command, obtain the actual output torque through actuator saturation limit, calculate the input saturation deviation, and apply the actual output torque to the drive wheelset; Step 4: Update the command filter using the projection trajectory, update the auxiliary dynamic system state using the projection deviation, input saturation deviation and recursive tracking error, and enter the next control cycle.

2. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 1, characterized in that, In step 1 of the controller construction, the wheel-rail adhesion dynamics model and the actuator input saturation model are respectively: (1) (2) wherein, , and are the train displacement, the running speed and the wheelset angular velocity, respectively, is the train mass, is the wheelset moment of inertia, is the viscous damping coefficient, , and are the wheel-rail adhesion force, the running resistance and the adhesion torque, respectively, , and are external disturbances and unmodeled dynamics, is the traction or braking torque command, is the actual output torque of the actuator, and are the lower and upper limits of the allowable output torque, respectively; the slip ratio is calculated as follows: (3) in, For sliding speed, For creep reference speed, The radius is the wheel radius.

3. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 1, characterized in that, In the controller construction step 2, with , , The state transition is performed, and the strict feedback form is a third-order strict feedback form: (4) in, , , , and These are train displacement, running speed, and slip speed, respectively. For system output, The nonlinear function determined for the model, To control the gain, This refers to the lumped uncertainty term formed by external disturbances, parameter uncertainties, and adhesion model errors. , For traction or braking torque command The actual output torque after input saturation limiting; the state constraints are set as follows: (5) (6) in, This is the upper limit of displacement. and These are the lower and upper limits of the running speed, respectively. To set the upper limit of the allowable slip velocity amplitude, To preset the upper limit of safe slip ratio, A small positive number used to construct the slip velocity boundary. The reference velocity used for calculating the slip velocity boundary.

4. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 1, characterized in that, In step 3 of the controller construction, the projection mapping is smooth, strictly monotonic, and invertible, its first derivative is bounded, and satisfies: (7) in, For time, and These are the ideal virtual control trajectories for the running speed and the sliding speed, respectively. For the corresponding projection trajectory, and These represent the lower and upper boundaries of the allowed interval for the corresponding state. This is to provide a safety margin for the projection.

5. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 4, characterized in that, The projection mapping adopts a hyperbolic tangent form: (8) (9) in, and These are the half-width and center of the projection interval after reserving a safety margin. These are the projection shape parameters.

6. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 1, characterized in that, In step 4 of the controller construction, the projection deviation and input saturation deviation are defined as follows: (10) in, and For two-level ideal virtual control trajectories, For the corresponding projection trajectory, For time, For traction or braking torque command, The actual output torque of the actuator; the auxiliary dynamic system is: (11) in, For the first Level 2 auxiliary dynamic system status, For the first Level recursive tracking error, This is the upper bound of the corresponding control gain. To attenuate the gain, Smoothing parameters representing the arctangent compensation term; two-stage projection bias , and input saturation deviation Enter separately , and The dynamic equations, and the subsequent auxiliary state Enter the previous level of auxiliary dynamics, This defines the terminal boundary.

7. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 6, characterized in that, In step 5 of the controller construction, the command filter, command filtering error, and recursive tracking error are respectively: (12) (13) (14) in, The time constant for the command filter. For the output of the command filter, Its time derivative, For command filtering error, For train displacement, For the desired displacement trajectory, and These are the running speed and the sliding speed, respectively; the filter output and its derivative are used to construct the subsequent robust backstepping control law.

8. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 7, characterized in that, In step 6 of the controller construction, an intermediate function is constructed and two levels of ideal virtual control trajectory and traction or braking torque command are generated sequentially: (15) (16) (17) (18) in, , and For intermediate functions, For nonlinear functions in a strict feedback model, The time derivative of the desired displacement trajectory. For error feedback gain, For robust smoothing parameters, and Control gain The lower and upper bounds, For lumped uncertainty terms The upper bound of the amplitude, and .

9. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 8, characterized in that, In step 7 of the controller construction, the command filter time constant and the attenuation gain of the auxiliary dynamic system satisfy the following: (19) in, , and The attenuation margin is a positive design constant; the projection safety margin satisfies: (20) in, and These represent the lower and upper boundaries of the allowed interval for the corresponding state. For projection safety margin, , and These are the upper bounds of the amplitudes of the command filtering error, recursive tracking error, and auxiliary dynamic system state, determined under the given initial conditions, over the entire control time domain. These upper bounds include the effects of the initial transient.

10. The high-speed train adhesion constraint balance control method based on virtual trajectory projection according to claim 1, characterized in that, In the online closed-loop control, the command filter state and auxiliary dynamic system state of the previous control cycle are read in each control cycle and updated in the following order: Calculate the first-level recursive tracking error based on the current train displacement, the desired displacement trajectory and the first auxiliary state, generate the ideal virtual control trajectory of the running speed, complete the first-level projection and calculate the first-level projection deviation; Calculate the second-level recursive tracking error based on the current running speed, the corresponding command filter output, and the second auxiliary state; generate the ideal virtual control trajectory of the slip speed; complete the second-level projection; and calculate the second-level projection deviation. The third-level recursive tracking error is calculated based on the current slip speed, the corresponding command filter output, and the third auxiliary state. A traction or braking torque command is generated, and the actual output torque is obtained after input saturation limiting. The input saturation deviation is also calculated. The two-stage projection trajectories are input into the corresponding command filters, and the two-stage projection deviations and input saturation deviations, along with the corresponding recursive tracking errors, are input into the auxiliary dynamic system. The state of the command filters and the auxiliary states at each stage are updated synchronously for use in the next control cycle.

Citation Information

Patent Citations

  • High-speed train optimal adhesion braking control method and system

    CN109484384A

  • A method for high-speed train adhesion and anti-slip control based on asymmetric Barrier Lyapunov function

    CN110095979B