Method and device for controlling dynamic uncoupling of hooks at the same speed in the train operation state

Abstract

This invention provides a method and apparatus for dynamic uncoupling and recoupling operations at common speed during train operation, relating to the field of uncoupling and recoupling control technology. The method includes: during the approach of the main control vehicle to the target car, performing linkage control calculations on the target car to obtain longitudinal following control quantities and pose adjustment control quantities; sensing local environmental information on the main control vehicle to obtain environmental constraint vectors, performing constraint optimization to obtain optimized longitudinal control quantities; monitoring relative pose to obtain a relative pose deviation sequence; performing sliding consistency judgment to trigger end-effector commands; and implementing force-position hybrid closed-loop control until the coupler mechanism reaches and locks into the target operating state. This invention solves the technical problems of existing technologies that mainly use fixed-distance triggering or single-speed following methods for uncoupling and recoupling operations, which cannot accurately reflect the actual coupling state of the coupler and are prone to coupler misalignment, collision impact, and uncoupling / recoupling failure.

Description

Technical Field

[0001] This invention relates to the field of uncoupling and recoupling control technology, specifically to a method and device for controlling dynamic uncoupling and recoupling operations at common speed during train operation. Background Technology

[0002] As railway transportation systems develop towards intelligence, automation, and high efficiency, the demand for automatic uncoupling and automatic recoupling technologies during train marshalling, demarcation, and vehicle transfer is constantly increasing. Traditional static uncoupling and recoupling methods relying on manual stopping are no longer sufficient to meet the demands of continuous and high-efficiency operations. Therefore, intelligent control technology that enables the main control vehicle and the target carriage to approach at the same speed and complete dynamic uncoupling and recoupling while the train is in motion is gradually becoming an important research direction in the field of intelligent control for rail transit.

[0003] In existing technologies, some automatic uncoupling and recoupling systems mainly use a fixed-distance triggering method or a single-speed following method for control. That is, after detecting the distance between vehicles through a ranging device, the system controls the working vehicle to approach the target car and performs the uncoupling and recoupling action when the preset distance is reached. However, during train operation, the target car is affected by track curves, gradient changes, wheel-rail vibrations, and vehicle serpentine movements, causing the coupler's spatial posture to continuously change. As a result, the traditional fixed-distance triggering method cannot accurately reflect the actual coupling state of the coupler, which can easily lead to problems such as coupler misalignment, collision impact, and uncoupling / recoupling failure. Summary of the Invention

[0004] This application provides a method and device for controlling the dynamic uncoupling and recoupling operations at common speed during train operation. It aims to solve the technical problems in the prior art, which mainly uses a fixed distance triggering method or a single speed following method for uncoupling and recoupling operations, which cannot accurately reflect the actual docking status of the coupler and is prone to coupler misalignment, collision impact, and uncoupling and recoupling failure.

[0005] The first aspect of this application discloses a common-speed dynamic uncoupling and recoupling operation control method under train operation conditions. The method includes: during the process of the main control operation vehicle approaching the target car, performing linkage control calculation based on state perception prediction on the target car to obtain longitudinal following control quantity and pose adjustment control quantity; performing local environmental information perception on the main control operation vehicle to obtain an environmental constraint vector, and performing constraint optimization of the longitudinal following control quantity to obtain an optimized longitudinal control quantity; during the closed-loop tracking control of the main control operation vehicle using the optimized longitudinal control quantity, performing relative pose monitoring through an on-board sensing network to obtain a relative pose deviation sequence; dynamically setting a multi-dimensional safety margin based on the environmental constraint vector, performing sliding consistency judgment of the relative pose deviation sequence, and triggering an end-effector command based on the pose adjustment control quantity; and implementing force-position hybrid closed-loop control during the process of controlling the pose adjustment mechanism of the main control operation vehicle to perform uncoupling and recoupling actions based on the end-effector command, until the coupler mechanism reaches and locks into the target operation state.

[0006] The second aspect of this application discloses a common-speed dynamic uncoupling and recoupling operation control device under train operation conditions. The device is used in the aforementioned common-speed dynamic uncoupling and recoupling operation control method under train operation conditions. The device includes: a linkage control calculation module, used to perform linkage control calculation based on state perception prediction on the target car during the approach of the main control operation vehicle to the target car, obtaining longitudinal following control quantity and pose adjustment control quantity; a constraint optimization module, used to perceive local environmental information of the main control operation vehicle, obtain an environmental constraint vector, and perform constraint optimization of the longitudinal following control quantity to obtain an optimized longitudinal control quantity; and a relative pose monitoring module. The measurement module is used to monitor the relative pose through the on-board sensing network and obtain the relative pose deviation sequence during the closed-loop tracking control of the main control vehicle using the optimized longitudinal control quantity; the consistency judgment module is used to dynamically set a multi-dimensional safety margin based on the environmental constraint vector, perform sliding consistency judgment on the relative pose deviation sequence, and trigger the end servo command based on the pose adjustment control quantity; the closed-loop control module is used to control the pose adjustment mechanism of the main control vehicle to perform the hook-and-coupler action based on the end servo command, and implement force-position hybrid closed-loop control until the coupler mechanism reaches and locks into the target working state.

[0007] One or more technical solutions provided in this application have at least the following beneficial effects: By using the main control vehicle to perform state perception prediction and linkage control calculations on the target car, the main control vehicle can obtain the longitudinal motion trend and coupler pose changes of the target car in the future time domain in advance. This allows for the synchronous generation of longitudinal following control and pose adjustment control quantities, achieving common speed approach and coordinated attitude adjustment between the main control vehicle and the target car. The main control vehicle also performs real-time perception of the track environment, wheel-rail adhesion state, and meteorological environment, and optimizes the longitudinal following control quantity based on environmental constraint vectors. This enables the traction and braking forces output by the main control vehicle to dynamically adapt to changes in the current operating environment, thereby improving the stability and safety of common speed following control. Furthermore, by continuously monitoring the relative pose of the target car during closed-loop tracking, the main control vehicle can obtain real-time changes in lateral deviation, vertical deviation, and attitude angle deviation between the two, forming a relative pose deviation sequence. This provides continuous and dynamic position data for subsequent uncoupling and recoupling operations. Based on the attitude criteria, the real-time performance and stability of the docking process under operating conditions are improved. By dynamically setting multi-dimensional safety margins based on environmental constraint vectors and performing sliding consistency judgment on the relative posture deviation sequence, the stability of the current docking state can be determined by comprehensively considering the line status, environmental disturbances, and posture fluctuations during operation. The end servo action is only triggered when the target car and the main control work car reach a continuous and stable matching state, thereby reducing the risk of mis-coupling or mis-re-coupling caused by instantaneous fluctuations or short-term misjudgments, and improving the safety and reliability of dynamic uncoupling and re-coupling operations. The main control work car controls the posture adjustment mechanism to execute the uncoupling and re-coupling actions, and implements force-position hybrid closed-loop control during the execution process. This allows the coupler mechanism to adjust the contact force and displacement changes in real time while maintaining precise posture alignment, reducing the risk of coupler collision impact and mechanical jamming, improving the smoothness and locking reliability of automatic coupler uncoupling and re-coupling operations under operating conditions, and thus improving the efficiency of dynamic train formation operations.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 This is a schematic flowchart of a common-speed dynamic uncoupling and recoupling operation control method provided in an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of the common-speed dynamic uncoupling and recoupling operation control device provided in the embodiment of this application.

[0011] Explanation of reference numerals in the attached diagram: Linkage control solution module 10, constraint optimization module 20, relative pose monitoring module 30, consistency judgment module 40, closed-loop control module 50. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown in the embodiment of this application, a method for controlling dynamic uncoupling and recoupling operations at common speed during train operation is provided. The method includes: During the process of the main control vehicle approaching the target carriage, the target carriage is subjected to linkage control calculation based on state perception prediction to obtain longitudinal following control quantity and pose adjustment control quantity.

[0014] During the approach of the main control vehicle to the target carriage, the main control vehicle continuously perceives the target carriage's state through the onboard sensing network, acquiring the target carriage's speed, acceleration, direction of travel, and coupler spatial pose information, and forming a corresponding carriage motion state sequence. The motion state sequence is then time-series predicted to obtain the longitudinal motion state and coupler pose state of the target carriage in the prediction time domain. Based on this, the main control vehicle, considering its own dynamic constraints, performs follow-up control calculations on the longitudinal motion state of the target carriage, obtaining longitudinal follow-up control quantities for controlling the main control vehicle's speed and traction / braking output; simultaneously, pose servo calculations are performed based on the coupler pose state of the target carriage, and feedforward compensation is applied to the vehicle body attitude disturbances generated during the main control vehicle's movement, obtaining pose adjustment control quantities for controlling the pose adjustment mechanism.

[0015] The main control vehicle is subjected to local environmental information perception to obtain an environmental constraint vector, and the longitudinal following control quantity is optimized to obtain an optimized longitudinal control quantity.

[0016] During the follow-up control process, the main control vehicle perceives the local operating environment in real time, acquiring digital information about the track, wheel-rail adhesion status, and meteorological conditions, and constructs an environmental constraint vector. Based on environmental factors such as track gradient, curve radius, and wind resistance, it calculates and predicts changes in external resistance over the time domain. Combining this with real-time wheel-rail adhesion capability, it dynamically corrects the range of traction and braking forces that the main control vehicle can output, forming a corresponding traction-braking feasible domain. The longitudinal follow-up control quantity is mapped to this traction-braking feasible domain for constraint optimization. Under the conditions of satisfying operational safety and vehicle dynamics constraints, the optimized longitudinal control quantity is output.

[0017] During the closed-loop tracking control of the main control vehicle using the optimized longitudinal control quantity, relative pose monitoring is performed through the on-board sensing network to obtain the relative pose deviation sequence.

[0018] The main control vehicle performs closed-loop tracking control based on optimized longitudinal control values, ensuring that it approaches the target car at the same speed. During the tracking process, the onboard sensing network continuously monitors the relative positional relationship between the main control vehicle and the target car, measuring in real time the lateral deviation, vertical deviation, and relative attitude angle deviation between the couplers, and forming a relative posture deviation sequence in chronological order for subsequent safety judgment and servo control of uncoupling and recoupling actions.

[0019] Based on the environmental constraint vector, a multi-dimensional safety margin is dynamically set, and the sliding consistency of the relative pose deviation sequence is judged, triggering an end-effector command based on the pose adjustment control quantity.

[0020] The main control vehicle dynamically sets multi-dimensional safety margins based on environmental constraint vectors. These margins include allowable lateral displacement, allowable vertical height, allowable relative attitude angle, and allowable deviation change rate. A sliding window statistical analysis is performed on the relative pose deviation sequence to extract corresponding multi-dimensional statistical features, and time-step consistency is determined based on the multi-dimensional safety margins. When the relative pose deviation continuously meets preset safety conditions, it is determined that the main control vehicle and the target carriage have reached a stable docking state, thereby triggering end-effector servo commands based on pose adjustment control quantities.

[0021] Based on the end-effector servo command, the position adjustment mechanism of the main control vehicle is controlled to perform the hook-and-couple action, and force-position hybrid closed-loop control is implemented until the coupler mechanism reaches and locks into the target working state.

[0022] After receiving the end-effector command, the main control vehicle controls the posture adjustment mechanism to perform uncoupling or recoupling actions. During the execution of the action, the posture adjustment mechanism drives the coupler to perform fine adjustments in the lateral, vertical, and attitude directions based on the posture adjustment control quantity, and combines the coupler posture information fed back in real time by the on-board sensing network to perform closed-loop correction of the posture adjustment process. At the same time, force-position hybrid closed-loop control is implemented on the force and displacement state during the coupler contact process to avoid rigid collisions or misalignment impacts, until the coupler mechanism completes the uncoupling or recoupling action and is stably locked in the target working state.

[0023] Furthermore, during the process of the main control vehicle approaching the target carriage, the target carriage undergoes linkage control calculation based on state perception prediction to obtain longitudinal following control and pose adjustment control quantities, including: Through the onboard perception network of the main control vehicle, during the process of the main control vehicle approaching the target carriage, multi-source synchronous state perception of the target carriage is performed to obtain the carriage motion state sequence; trajectory attitude prediction based on time extrapolation is performed on the carriage motion state sequence to obtain the longitudinal motion state and coupler pose state; linkage control calculation is performed on the longitudinal motion state and coupler pose state to obtain the longitudinal following control quantity and pose adjustment control quantity.

[0024] As the main control vehicle approaches the target carriage, it uses an onboard sensing network to perform multi-source synchronous state perception of the target carriage. This onboard sensing network includes visual sensors, millimeter-wave radar, a laser rangefinder, an inertial measurement unit, and an onboard communication module, used to acquire real-time information on the target carriage's speed, acceleration, position changes, and coupler attitude. The various sensing data are then synchronized in time, unified in coordinates, and fused to form a sequence of carriage motion states reflecting changes in the target carriage's operating status.

[0025] Trajectory and attitude prediction based on temporal extrapolation is performed on the motion state sequence of the target car. First, the longitudinal running data of the target car is analyzed to predict the velocity and acceleration changes of the target car in the subsequent prediction time domain, thus obtaining the longitudinal motion state. At the same time, the lateral offset, vertical fluctuation, and attitude angle changes of the coupler of the target car are dynamically predicted, and the position and attitude change relationship is established by combining the digital information of the track curve, gradient, etc., to obtain the coupler position and attitude state of the target car.

[0026] The linkage control is calculated based on the predicted longitudinal motion state and coupler pose state. First, the following trajectory of the main vehicle is planned according to the longitudinal motion trend of the target car, and the longitudinal following control quantity is solved in combination with the dynamic constraints of the main vehicle to control the traction and braking output of the main control vehicle. At the same time, according to the spatial pose state of the target car coupler, the pose adjustment mechanism of the main control vehicle is servo controlled and compensated for the attitude disturbance generated during the operation of the main control vehicle to obtain the pose adjustment control quantity used to control the coupler docking attitude.

[0027] Furthermore, trajectory and attitude prediction based on temporal extrapolation is performed on the motion state sequence of the carriage to obtain the longitudinal motion state and coupler pose state, including: The motion state sequence of the carriage is decoupled and separated to obtain the macroscopic motion sequence of the carriage and the microscopic motion sequence of the coupler. The macroscopic motion sequence of the carriage is optimally estimated to obtain the optimal motion state vector. Then, key motion parameters are identified to obtain the longitudinal kinematic parameters. Constrained forward prediction is performed by combining the digital information of the track and the longitudinal kinematic parameters to output the longitudinal motion state. The pose fluctuation coupled dynamics modeling of the microscopic motion sequence of the coupler is performed based on the digital information of the track to output the pose state of the coupler.

[0028] The motion state sequence of the train car is decoupled based on its motion characteristics. Since the longitudinal velocity changes generated by the overall movement of the target train car and the local oscillations of the coupler have different frequency characteristics, the motion state sequence of the train car is processed by layering and separating based on the correlation between motion amplitude, frequency of change, and spatial position. Specifically, low-frequency, high-amplitude overall operating data is identified as the macroscopic motion sequence of the train car, used to characterize the overall longitudinal operating state of the target train car; high-frequency, low-amplitude local fluctuation data is identified as the microscopic motion sequence of the coupler, used to characterize the lateral oscillation, vertical jump, and attitude angle changes of the coupler during operation, thereby achieving separate modeling of the overall motion of the target train car and the local motion of the coupler.

[0029] Optimal state estimation is performed on the macroscopic motion sequence of the train carriage. Using velocity, acceleration, and displacement change data at consecutive time points, the current operating state of the target carriage is filtered and fused to reduce random disturbances caused by perceived noise and track vibration, resulting in a stable optimal motion state vector. Based on this optimal motion state vector, key motion parameters of the target carriage are identified, yielding its longitudinal kinematic parameters under the current operating conditions. These longitudinal kinematic parameters include operating velocity, acceleration, rate of change of velocity, and the trend of change in operating inertia, used to describe the longitudinal motion characteristics of the target carriage in subsequent time domains.

[0030] By combining digital information about the railway line and longitudinal kinematic parameters, a constrained forward prediction is performed on the target carriage. First, based on the current operating position, the curve radius, gradient changes, and direction information of the corresponding line segment within the prediction time domain are extracted from the digital information. Then, the additional running resistance of the target carriage during subsequent operation is calculated based on the line parameters. Using the longitudinal kinematic parameters as the initial input to the vehicle's longitudinal dynamics model, and combining the target carriage's inertial mass, running resistance, and braking state, a rolling time-domain forward simulation is performed to predict the velocity and acceleration trends of the target carriage in the future time domain, thereby outputting the longitudinal motion state.

[0031] A coupled dynamic model of pose fluctuations is performed on the micro-motion sequence of the coupler. Since the car body vibration is transmitted to the coupler structure when the target car is running on curved tracks, slopes, and under uneven track conditions, the main control vehicle combines digital track information to perform coupled analysis of lateral offset, vertical fluctuations, and attitude angle changes of the coupler. Based on the coupler's micro-motion sequence, a coupler pose fluctuation model is established. Combining changes in track curvature, gradient, and vehicle vibration transmission, the spatial pose changes of the coupler in the prediction time domain are dynamically predicted, and the corresponding coupler pose state is output for subsequent pose servo control during the uncoupling and recoupling process.

[0032] Furthermore, by combining the digital information of the line and the longitudinal kinematic parameters to perform constrained forward prediction, the longitudinal motion state is output, including: The current geodetic coordinates of the main control vehicle are extracted. The track curve radius sequence and slope angle sequence for the prediction time domain are extracted from the digital information of the line. The additional resistance model of the line is calculated. Based on the equivalent inertial mass of the target car, the basic running resistance parameters, and the additional resistance model of the line, a longitudinal dynamics model of the vehicle is constructed. The longitudinal kinematic parameters are used as the initial state of the longitudinal dynamics model of the vehicle. Combined with the equivalent braking resistance parameters, a forward rolling time domain simulation is performed to output the longitudinal motion state, wherein the longitudinal motion state includes a velocity prediction sequence and an acceleration prediction sequence.

[0033] The main control vehicle acquires the geodetic coordinates of its current operating position and matches the current track segment in the track digital information based on these coordinates. It extracts the track curve radius sequence and gradient angle sequence for the corresponding track within the prediction time domain from the track digital information, and calculates the additional running resistance experienced by the target car during subsequent operation based on changes in track curvature and gradient. Specifically, the curve radius change characterizes the additional curve resistance generated by the target car when running on a curved track, and the gradient angle change characterizes the gradient resistance generated by the target car when running on a slope. The curve resistance and gradient resistance are comprehensively calculated to form a track additional resistance model within the corresponding prediction time domain, used to describe the impact of track conditions on the longitudinal running state of the target car.

[0034] The main control vehicle constructs a longitudinal dynamics model of the vehicle based on the equivalent inertial mass of the target car, basic operating resistance parameters, and the track-addressed resistance model. The equivalent inertial mass parameters of the target car are determined according to its formation mass, load state, and operating inertial characteristics. A basic operating resistance parameter model is established by combining the mechanical resistance, air resistance, and wheel-rail running resistance experienced by the target car during operation. The track-addressed resistance model is introduced into the vehicle's longitudinal dynamics model, enabling it to simultaneously reflect the target car's own operating characteristics and the impact of track environment changes on the vehicle's operating state, thus forming the longitudinal dynamic constraints of the target car under actual operating conditions.

[0035] The longitudinal kinematic parameters are used as the initial state input to the vehicle's longitudinal dynamics model. Combined with the equivalent braking resistance parameters of the target carriage, a forward rolling time-domain simulation of the target carriage's subsequent operating state is performed. During the simulation, the main control vehicle continuously iterates and calculates the future operating state of the target carriage at a preset time step. At each simulation moment, the target carriage's speed and acceleration are updated based on the vehicle's longitudinal dynamics model, and the impact of changes in the track's additional resistance on the target carriage's operating state is simultaneously superimposed. The iterative results throughout the entire prediction time domain are output in a time sequence to obtain the longitudinal motion state of the target carriage. This longitudinal motion state includes the target carriage's predicted speed and acceleration sequences within the prediction time domain, which are used by the main control vehicle to execute common-speed following control.

[0036] Furthermore, the longitudinal motion state and the coupler pose state are subjected to linkage control calculation to obtain the longitudinal following control quantity and the pose adjustment control quantity, including: Based on the following safety distance and the longitudinal motion state, the desired trajectory of the main control vehicle is planned to obtain the longitudinal reference trajectory of the vehicle. Under the dynamic constraints of the main control vehicle, rolling time-domain optimization control is performed based on the longitudinal reference trajectory to obtain the longitudinal following control quantity. Pose servo calculation is performed based on the coupler pose state to output the initial pose control quantity. The longitudinal following control quantity is used to perform forward kinematics simulation of the main control vehicle to predict the coupling effect of the vehicle body attitude change on the coupler pose, and an additional pose disturbance is output. Feedforward compensation of the initial pose control quantity is performed based on the additional pose disturbance to obtain the pose adjustment control quantity.

[0037] The main control vehicle plans its desired trajectory based on the longitudinal motion state of the target carriage. First, the main control vehicle reads the target carriage's predicted speed sequence, acceleration sequence, and position change information within the prediction time domain. Combined with a preset following safety distance, it determines the allowable target distance range between the main control vehicle and the target carriage. Then, based on the target carriage's future movement trend, the main control vehicle continuously plans its own speed, following distance, and acceleration / deceleration changes within the prediction time domain. This allows the main control vehicle to gradually approach the target carriage at the same speed while maintaining a safe distance, thus generating a longitudinal reference trajectory for the vehicle.

[0038] Under its own dynamic constraints, the main control vehicle performs rolling time-domain optimization control based on its longitudinal reference trajectory. The main control vehicle establishes its own longitudinal dynamic model, using vehicle traction capacity, braking capacity, running inertia, and speed variation limits as control constraints. Within each control cycle, the main control vehicle performs rolling optimization calculations on the control inputs in the future time domain based on the current operating state and the predicted reference trajectory, ensuring that the actual operating trajectory of the main control vehicle continuously approximates the longitudinal reference trajectory, and outputs the corresponding longitudinal following control quantity. This longitudinal following control quantity is used to control the traction or braking output of the main control vehicle.

[0039] The main control vehicle acquires the lateral position, vertical height, and attitude angle changes of the target car coupler in the prediction time domain, and calculates the pose deviation between the two by combining the current position of the main control vehicle's own coupler. Based on the pose deviation, the main control vehicle performs servo solving on the adjustment direction and adjustment amount of the pose adjustment mechanism to obtain the initial pose control quantities used to control the lateral movement, vertical adjustment, and attitude correction of the coupler.

[0040] Forward kinematic simulation is performed using longitudinal following control variables to predict the coupling effect of changes in the main control vehicle's body attitude on the coupler's pose. Since the main control vehicle experiences pitch, yaw, and vibration changes during acceleration, deceleration, and curve movement, the future operating attitude of the main control vehicle is dynamically simulated based on the longitudinal following control variables, and the transmission relationship between body attitude changes and the coupler structure is analyzed. The additional effects of body attitude changes on the coupler's lateral position, vertical height, and attitude angle are calculated, thus outputting the additional pose disturbance.

[0041] The main control vehicle first maps the additional posture disturbance to the corresponding coupler adjustment direction, and pre-corrects the initial posture control quantity based on the disturbance change trend, so that the posture adjustment mechanism can complete the pre-compensation adjustment before the posture of the main control vehicle changes. The compensated posture adjustment control quantity is output to reduce the impact of vehicle movement on coupler docking accuracy and improve the stability and docking accuracy of the uncoupling and recoupling process under operating conditions.

[0042] Furthermore, the environmental constraint vector includes line digital information, adhesion estimation information, and meteorological environmental information.

[0043] The track digital information reflects the track structure status of the current operating section of the main control vehicle, including track gradient, curve radius, track direction, and speed limit information. Adhesion estimation information reflects the adhesion capability under current wheel-rail contact conditions, including wheel-rail adhesion coefficient, wheel-rail slippage state, and wheelset adhesion stability. Meteorological environment information reflects the external meteorological conditions in the current operating environment of the main control vehicle, including wind speed, wind direction, rainfall, and ambient temperature. The main control vehicle collects and fuses the above information in real time through an onboard sensing network to form an environmental constraint vector for longitudinal control constraint optimization.

[0044] Furthermore, the main control vehicle is subjected to local environmental information perception to obtain an environmental constraint vector, and the longitudinal following control quantity is constrained and optimized to obtain an optimized longitudinal control quantity, including: The gradient resistance component and curve resistance component are extracted from the line additional resistance model; the real-time wheel-rail adhesion coefficient is extracted from the adhesion estimation information, and the available traction force limit and available braking force limit are calculated in combination with the axle load distribution of the main control vehicle; the real-time wind speed and real-time wind direction are extracted from the meteorological environment information, and the wind resistance component of the main control vehicle on the windward side is calculated; the gradient resistance component, curve resistance component, and wind resistance component are vector synthesized to output the external resistance prediction sequence within the prediction time domain interval; the external resistance prediction sequence is used as the dynamic compensation benchmark to perform dynamic boundary correction of the available traction force limit and available braking force limit, and the traction-braking force feasible domain is output; the longitudinal following control quantity is projected onto the traction-braking force feasible domain to solve for minimizing the tracking error, and the optimized longitudinal control quantity is output.

[0045] Based on the current operating position, the slope angle change of the corresponding line section in the prediction time domain is matched and the slope resistance change value is calculated based on the vehicle gravity component; based on the change of the line curve radius and the vehicle curve operation characteristics, the curve additional resistance generated when the vehicle passes through the curve is calculated, thereby obtaining the slope resistance component and curve additional resistance component in the prediction time domain.

[0046] Based on the wheelset speed difference, wheel-rail slippage state, and wheel-rail contact stability, the current wheel-rail adhesion capacity is estimated to obtain the real-time wheel-rail adhesion coefficient. Combined with the axle load distribution state of each wheelset of the main control vehicle, the maximum traction force and maximum braking force that can be stably transmitted between the wheels and rails under the current working conditions are calculated to prevent the main control vehicle from spinning or sliding during longitudinal following, thereby obtaining the corresponding available traction force limit and available braking force limit.

[0047] The system acquires information on wind speed and direction changes in the current operating environment. Combined with the current operating direction of the main control vehicle, it calculates the angle of attack of the airflow relative to the vehicle body. Based on the windward area of ​​the main control vehicle and the vehicle's aerodynamic parameters, it calculates the changes in air resistance experienced by the vehicle during operation, thereby obtaining the wind resistance component for subsequent dynamic constraint optimization of longitudinal following control quantities.

[0048] Based on the direction of action of each resistance component, a unified directional mapping is performed on the operating resistance from different sources. Furthermore, the resistance changes are time-series corresponded according to the line and environmental conditions at each moment within the prediction time domain. Gradient resistance, curve-related resistance, and wind resistance are superimposed and calculated to obtain the comprehensive external resistance value corresponding to each time step within the prediction time domain, thus forming an external resistance prediction sequence. This sequence is used to characterize the dynamic external load changes experienced by the main control vehicle during subsequent operation.

[0049] Based on the changes in comprehensive external resistance within the predicted time domain, the actual traction and braking outputs required by the vehicle are dynamically estimated. Combining real-time wheel-rail adhesion capacity and the current axle load distribution, the original available traction and braking force limits are narrowed or adjusted to prevent traction overload, insufficient braking, or wheel-rail slippage of the main control vehicle on complex tracks or in low-adhesion environments. Based on the corrected traction and braking capacity ranges, a feasible traction-braking force domain that meets the current environmental constraints is constructed.

[0050] Determine whether the current longitudinal following control quantity exceeds the boundary of the traction-braking force feasible domain. When the control quantity exceeds the feasible range, perform boundary constraint mapping on the control quantity. Take the minimum tracking error of the main control vehicle on the longitudinal reference trajectory as the optimization objective. Under the condition of satisfying the traction-braking force feasible domain constraint, optimize and adjust the control output so that the main control vehicle can maintain stable tracking capability while meeting the dynamic safety requirements under the current environmental conditions. Finally, output the optimized longitudinal control quantity.

[0051] Furthermore, based on the environmental constraint vector, a multi-dimensional safety margin is dynamically set, and a sliding consistency judgment is performed on the relative pose deviation sequence to trigger an end-effector command based on the pose adjustment control quantity, including: The environmental constraint vector is mapped using multiple parameters to obtain the multidimensional safety margin, which includes a lateral displacement threshold, a vertical height threshold, a relative attitude angle threshold, and a deviation change rate threshold. Multidimensional feature sliding statistics are performed on the relative pose deviation sequence to obtain a multidimensional statistical feature sequence. The multidimensional safety margin is used to map and traverse the multidimensional statistical feature sequence, performing a time-step consistency decision and outputting the decision result. If the decision passes, an end-effector command based on the pose adjustment control quantity is triggered.

[0052] The system reads digital information about the track, wheel-rail adhesion status, and meteorological conditions, and analyzes the impact of the current operating environment on the stability of the coupling / uncoupling process. Specifically, when the track curvature increases, the gradient changes more significantly, or the wheel-rail adhesion capacity decreases, the allowable positional deviation range of the main control vehicle is correspondingly reduced; when environmental stability is high, the corresponding positional tolerance range is appropriately widened. Based on the mapping relationship between different environmental parameters and the stability of the coupling / uncoupling, lateral displacement thresholds, vertical height thresholds, relative attitude angle thresholds, and deviation change rate thresholds are dynamically generated, thus forming a multi-dimensional safety margin for determining the safety of coupling / uncoupling.

[0053] The vehicle-mounted sensing network continuously acquires data on lateral deviation, vertical deviation, and attitude angle deviation between the main control vehicle and the target carriage, forming a relative pose deviation sequence in chronological order. Using a sliding time window approach, statistical analysis is performed on the pose deviation changes over consecutive time periods, extracting corresponding statistical features such as mean change, fluctuation amplitude, rate of change, and stable duration. This forms a multi-dimensional statistical feature sequence for consistency determination, avoiding the influence of instantaneous vibrations or random disturbances on the hook-and-unhook judgment results.

[0054] The statistical characteristics corresponding to each time step are compared item by item with the lateral displacement threshold, vertical height threshold, relative attitude angle threshold, and deviation change rate threshold under the current environmental conditions, and it is determined whether each statistical characteristic remains within the allowable range. A consistency analysis is performed on the judgment results of multiple consecutive time steps. When the relative pose deviation meets the stability condition within a continuous time period, it is determined that a stable docking state has been formed between the current main control vehicle and the target carriage, and the corresponding decision result is output.

[0055] When the arbitration result meets the preset consistency conditions, the main control vehicle triggers an end-effector command based on the pose adjustment control quantity. Specifically, according to the current relative pose state of the coupler, the corresponding pose adjustment control quantity is invoked, and an end-effector control command is sent to the pose adjustment mechanism to drive the coupler to perform lateral alignment, vertical height correction, and attitude angle adjustment actions, so that the main control vehicle coupler gradually approaches the target docking position of the target car coupler, thereby establishing stable pose conditions for subsequent uncoupling or recoupling actions.

[0056] Furthermore, during the process of controlling the pose adjustment mechanism to perform the uncoupling and recoupling action based on the end-effector servo command, the pose adjustment control quantity is corrected in real time using the coupler pose information fed back by the vehicle-mounted sensing network.

[0057] During the uncoupling and recoupling actions of the posture adjustment mechanism controlled by the end-effector servo command, the main control vehicle first continuously acquires information on the relative position changes, attitude changes, and contact state changes between the couplers, and calculates the current coupler posture deviation in real time. Based on the real-time posture deviation, the posture adjustment control quantity is dynamically corrected, enabling the posture adjustment mechanism to continuously correct changes in the coupler's lateral position, vertical height, and attitude angle to compensate for the disturbances caused by vehicle vibration, track irregularities, and changes in vehicle posture during the coupler docking process. Simultaneously, during coupler contact, the main control vehicle performs joint closed-loop control on the coupler contact force and displacement changes to avoid rigid impacts, misalignment collisions, or locking failures, until the coupler mechanism completes the uncoupling or recoupling action and stably locks into the target working state.

[0058] Example 2 is based on the same inventive concept as the common-speed dynamic uncoupling and recoupling operation control method under the train running state in the previous examples, such as... Figure 2 As shown in the embodiment of this application, a common-speed dynamic uncoupling and recoupling operation control device is provided under train operation conditions. The device includes: The linkage control calculation module 10 is used to perform linkage control calculation based on state perception prediction on the target car during the process of the main control vehicle approaching the target car, and obtain the longitudinal following control quantity and the pose adjustment control quantity; the constraint optimization module 20 is used to perceive the local environment information of the main control vehicle, obtain the environmental constraint vector, and perform constraint optimization on the longitudinal following control quantity to obtain the optimized longitudinal control quantity; the relative pose monitoring module 30 is used to monitor the relative pose through the vehicle-mounted sensing network during the closed-loop tracking control of the main control vehicle using the optimized longitudinal control quantity, and obtain the relative pose deviation sequence; the consistency judgment module 40 is used to dynamically set a multi-dimensional safety margin according to the environmental constraint vector, perform sliding consistency judgment on the relative pose deviation sequence, and trigger the end servo command based on the pose adjustment control quantity; the closed-loop control module 50 is used to implement force-position hybrid closed-loop control during the process of the pose adjustment mechanism of the main control vehicle performing the uncoupling and recoupling action based on the end servo command, until the coupler mechanism reaches and locks into the target working state.

[0059] Furthermore, the linkage control calculation module 10 is used to perform the following operation steps: Through the onboard perception network of the main control vehicle, during the process of the main control vehicle approaching the target carriage, multi-source synchronous state perception of the target carriage is performed to obtain the carriage motion state sequence; trajectory attitude prediction based on time extrapolation is performed on the carriage motion state sequence to obtain the longitudinal motion state and coupler pose state; linkage control calculation is performed on the longitudinal motion state and coupler pose state to obtain the longitudinal following control quantity and pose adjustment control quantity.

[0060] Furthermore, the linkage control calculation module 10 is used to perform the following operation steps: The motion state sequence of the carriage is decoupled and separated to obtain the macroscopic motion sequence of the carriage and the microscopic motion sequence of the coupler. The macroscopic motion sequence of the carriage is optimally estimated to obtain the optimal motion state vector. Then, key motion parameters are identified to obtain the longitudinal kinematic parameters. Constrained forward prediction is performed by combining the digital information of the track and the longitudinal kinematic parameters to output the longitudinal motion state. The pose fluctuation coupled dynamics modeling of the microscopic motion sequence of the coupler is performed based on the digital information of the track to output the pose state of the coupler.

[0061] Furthermore, the linkage control calculation module 10 is used to perform the following operation steps: The current geodetic coordinates of the main control vehicle are extracted. The track curve radius sequence and slope angle sequence for the prediction time domain are extracted from the digital information of the line. The additional resistance model of the line is calculated. Based on the equivalent inertial mass of the target car, the basic running resistance parameters, and the additional resistance model of the line, a longitudinal dynamics model of the vehicle is constructed. The longitudinal kinematic parameters are used as the initial state of the longitudinal dynamics model of the vehicle. Combined with the equivalent braking resistance parameters, a forward rolling time domain simulation is performed to output the longitudinal motion state, wherein the longitudinal motion state includes a velocity prediction sequence and an acceleration prediction sequence.

[0062] Furthermore, the linkage control calculation module 10 is used to perform the following operation steps: Based on the following safety distance and the longitudinal motion state, the desired trajectory of the main control vehicle is planned to obtain the longitudinal reference trajectory of the vehicle. Under the dynamic constraints of the main control vehicle, rolling time-domain optimization control is performed based on the longitudinal reference trajectory to obtain the longitudinal following control quantity. Pose servo calculation is performed based on the coupler pose state to output the initial pose control quantity. The longitudinal following control quantity is used to perform forward kinematics simulation of the main control vehicle to predict the coupling effect of the vehicle body attitude change on the coupler pose, and an additional pose disturbance is output. Feedforward compensation of the initial pose control quantity is performed based on the additional pose disturbance to obtain the pose adjustment control quantity.

[0063] Furthermore, the environmental constraint vector includes line digital information, adhesion estimation information, and meteorological environmental information.

[0064] Furthermore, the constraint optimization module 20 is used to perform the following operation steps: The gradient resistance component and curve resistance component are extracted from the line additional resistance model; the real-time wheel-rail adhesion coefficient is extracted from the adhesion estimation information, and the available traction force limit and available braking force limit are calculated in combination with the axle load distribution of the main control vehicle; the real-time wind speed and real-time wind direction are extracted from the meteorological environment information, and the wind resistance component of the main control vehicle on the windward side is calculated; the gradient resistance component, curve resistance component, and wind resistance component are vector synthesized to output the external resistance prediction sequence within the prediction time domain interval; the external resistance prediction sequence is used as the dynamic compensation benchmark to perform dynamic boundary correction of the available traction force limit and available braking force limit, and the traction-braking force feasible domain is output; the longitudinal following control quantity is projected onto the traction-braking force feasible domain to solve for minimizing the tracking error, and the optimized longitudinal control quantity is output.

[0065] Furthermore, the consistency judgment module 40 is used to perform the following operation steps: The environmental constraint vector is mapped using multiple parameters to obtain the multidimensional safety margin, which includes a lateral displacement threshold, a vertical height threshold, a relative attitude angle threshold, and a deviation change rate threshold. Multidimensional feature sliding statistics are performed on the relative pose deviation sequence to obtain a multidimensional statistical feature sequence. The multidimensional safety margin is used to map and traverse the multidimensional statistical feature sequence, performing a time-step consistency decision and outputting the decision result. If the decision passes, an end-effector command based on the pose adjustment control quantity is triggered.

[0066] Furthermore, during the process of controlling the pose adjustment mechanism to perform the uncoupling and recoupling action based on the end-effector servo command, the pose adjustment control quantity is corrected in real time using the coupler pose information fed back by the vehicle-mounted sensing network.

[0067] Through the foregoing detailed description of the common speed dynamic uncoupling and recoupling operation control method under train operation conditions, those skilled in the art can clearly understand the common speed dynamic uncoupling and recoupling operation control device under train operation conditions in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section description.

[0068] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for controlling dynamic uncoupling and recoupling operations at common speed during train operation, characterized in that, The method includes: During the process of the main control vehicle approaching the target carriage, the target carriage is subjected to linkage control calculation based on state perception prediction to obtain longitudinal following control quantity and pose adjustment control quantity. The main control vehicle is subjected to local environmental information perception to obtain an environmental constraint vector, and the longitudinal following control quantity is optimized to obtain an optimized longitudinal control quantity. During the closed-loop tracking control of the main control vehicle using the optimized longitudinal control quantity, relative pose monitoring is performed through the on-board sensing network to obtain the relative pose deviation sequence; Based on the environmental constraint vector, a multi-dimensional safety margin is dynamically set, and the sliding consistency of the relative pose deviation sequence is judged to trigger the end-effector command based on the pose adjustment control quantity. Based on the end-effector servo command, the position adjustment mechanism of the main control vehicle is controlled to perform the hook-and-couple action, and force-position hybrid closed-loop control is implemented until the coupler mechanism reaches and locks into the target working state.

2. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 1, characterized in that, During the approach of the main control vehicle to the target carriage, the target carriage is subjected to linkage control calculation based on state perception prediction to obtain longitudinal following control quantity and pose adjustment control quantity, including: Through the on-board sensing network of the main control vehicle, the multi-source synchronous state sensing of the target carriage is performed as the main control vehicle approaches the target carriage, and the carriage motion state sequence is obtained. The trajectory and attitude prediction based on temporal extrapolation is performed on the motion state sequence of the carriage to obtain the longitudinal motion state and the coupler pose state; The longitudinal motion state and the coupler position state are subjected to linkage control calculation to obtain the longitudinal following control quantity and the position adjustment control quantity.

3. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 2, characterized in that, Trajectory and attitude prediction based on temporal extrapolation are performed on the motion state sequence of the carriage to obtain the longitudinal motion state and coupler pose state, including: By decoupling and separating the motion state sequence of the carriage, we obtain the macroscopic motion sequence of the carriage and the microscopic motion sequence of the coupler; After performing optimal state estimation on the macroscopic motion sequence of the carriage to obtain the optimal motion state vector, key motion parameters are identified to obtain longitudinal kinematic parameters. By combining the digital information of the line and the longitudinal kinematic parameters, constrained forward prediction is performed, and the longitudinal motion state is output. The micro motion sequence of the coupler is modeled using pose fluctuation coupled dynamics based on the digital information of the line, and the pose state of the coupler is output.

4. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 3, characterized in that, By combining the digital information of the line and the longitudinal kinematic parameters, constrained forward prediction is performed, and the longitudinal motion state is output, including: Extract the current geodetic coordinates of the main control vehicle, extract the track curve radius sequence and slope angle sequence of the prediction time domain from the digital information of the line, and calculate the additional resistance model of the line; Based on the equivalent inertial mass of the target carriage, the basic operating resistance parameters, and the additional resistance model of the line, a longitudinal dynamics model of the vehicle is constructed. Using the longitudinal kinematic parameters as the initial state of the vehicle's longitudinal dynamics model, and combining them with the equivalent braking resistance parameters, a forward roll time-domain simulation is performed to output the longitudinal motion state, which includes a velocity prediction sequence and an acceleration prediction sequence.

5. The method for controlling dynamic uncoupling and recoupling operations at common speed during train operation as described in claim 2, characterized in that, The longitudinal motion state and the coupler pose state are subjected to linkage control calculation to obtain the longitudinal following control quantity and the pose adjustment control quantity, including: Based on the following safety distance and the longitudinal motion state, the desired trajectory of the main control vehicle is planned to obtain the longitudinal reference trajectory of the vehicle. Under the dynamic constraints of the main control vehicle, the longitudinal following control quantity is obtained by solving the rolling time-domain optimization control based on the longitudinal reference trajectory of the vehicle. Based on the coupler pose state, pose servo calculation is performed, and the initial pose control quantity is output. The longitudinal following control quantity is used to perform forward kinematics simulation of the main control vehicle to predict the coupling effect of the vehicle body attitude change on the coupler pose and output additional pose disturbance. The initial pose control amount is fed forward to compensate for the additional pose disturbance, and the pose adjustment control amount is obtained.

6. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 4, characterized in that, The environmental constraint vector includes line digital information, adhesion estimation information, and meteorological environmental information.

7. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 6, characterized in that, The main control vehicle performs local environmental information perception to obtain an environmental constraint vector, and performs constraint optimization of the longitudinal following control quantity to obtain an optimized longitudinal control quantity, including: Extract the gradient resistance component and the curve resistance component from the aforementioned line additional resistance model; The real-time wheel-rail adhesion coefficient is extracted from the adhesion estimation information, and combined with the axle load distribution of the main control vehicle, the available traction force limit and available braking force limit are calculated. Real-time wind speed and direction are extracted from the meteorological environment information, and the wind resistance component of the main control vehicle on the windward side is calculated. The slope resistance component, curve-additional resistance component, and wind resistance component are vector-synthesized to output the external resistance prediction sequence within the prediction time domain interval. Using the external resistance prediction sequence as a dynamic compensation benchmark, the dynamic boundary correction of the available traction force limit and available braking force limit is performed, and the traction-braking force feasible domain is output. The longitudinal following control quantity is projected onto the traction-braking feasible region to minimize the tracking error, and the optimized longitudinal control quantity is output.

8. The method for controlling dynamic uncoupling and recoupling operations at common speed during train operation as described in claim 1, characterized in that, Based on the environmental constraint vector, a multi-dimensional safety margin is dynamically set, and a sliding consistency judgment is performed on the relative pose deviation sequence to trigger an end-effector command based on the pose adjustment control quantity, including: The environmental constraint vector is mapped using multiple parameters to obtain the multidimensional safety margin, wherein the multidimensional safety margin includes a lateral displacement threshold, a vertical height threshold, a relative attitude angle threshold, and a deviation change rate threshold. Multidimensional feature sliding statistics are performed on the relative pose deviation sequence to obtain a multidimensional statistical feature sequence; The multidimensional safety margin mapping is used to traverse the multidimensional statistical feature sequence, and a time-step consistency decision is made to output the decision result. If the ruling is successful, an end-effector command based on the pose adjustment control amount is triggered.

9. The common-speed dynamic uncoupling and recoupling operation control method under train operation status as described in claim 1, characterized in that, During the process of controlling the pose adjustment mechanism to perform the uncoupling and recoupling action based on the end servo command, the pose adjustment control quantity is corrected in real time in a closed loop based on the coupler pose information fed back in real time by the vehicle-mounted sensing network.

10. A common-speed dynamic uncoupling and recoupling operation control device during train operation, characterized in that, For implementing the common-speed dynamic uncoupling and recoupling operation control method under any one of claims 1-9, the apparatus comprises: The linkage control calculation module is used to perform linkage control calculation on the target car based on state perception prediction during the process of the main control vehicle approaching the target car, so as to obtain the longitudinal following control quantity and the pose adjustment control quantity. The constraint optimization module is used to perceive local environmental information of the main control vehicle, obtain environmental constraint vectors, perform constraint optimization of the longitudinal following control quantity, and obtain optimized longitudinal control quantity. The relative pose monitoring module is used to monitor the relative pose through the on-board sensing network and obtain the relative pose deviation sequence during the closed-loop tracking control of the main control vehicle using the optimized longitudinal control quantity. The consistency judgment module is used to dynamically set a multi-dimensional safety margin based on the environmental constraint vector, perform a sliding consistency judgment on the relative pose deviation sequence, and trigger an end-effector command based on the pose adjustment control quantity. The closed-loop control module is used to control the posture adjustment mechanism of the main control vehicle to perform the hook-and-coupler action based on the end servo command, and to implement force-position hybrid closed-loop control until the coupler mechanism reaches and locks into the target working state.

Images