Method and system for coordinated control of automatic coupling and uncoupling of car couplings

Abstract

This invention discloses a method and system for coordinated action control of automatic coupler uncoupling and recoupling, relating to the field of coupler coordinated control technology. The method includes: generating a target motion map based on coupler operating status information and disassembly actions during uncoupling and recoupling; generating a spatiotemporal constraint action set; constructing a virtual coupling phase axis, performing coordinated control analysis, and determining a coordinated lockable state interval; and synchronously adjusting the action execution rhythm of the master and slave robotic arms according to the coordinated lockable state interval, so that the couplers at both ends complete the uncoupling or recoupling operation according to the spatiotemporal constraint action set. This invention solves the technical problems of poor motion coordination in coupler uncoupling and recoupling operations, susceptibility to vehicle body disturbances and cycle conflicts, high jamming failure rate, and low efficiency of automated operations in existing technologies. It achieves precise coordination between the master and slave robotic arms and the vehicle body movement and the tipper cycle, improving the reliability and efficiency of automatic coupler uncoupling and recoupling operations.

Description

Technical Field

[0001] This invention relates to the field of coupler coordination control technology, specifically to a method and system for automatic coupler disengagement and re-engagement coordination control. Background Technology

[0002] Automatic uncoupling and recoupling of car couplers is a crucial link in connecting shunting, tipping, and the transfer of loaded and empty cars. Current coupler uncoupling and recoupling operations mostly rely on manual operation or single-arm independent execution. This presents problems such as the car body easily shifting position and experiencing dynamic disturbances during shunting, and the difficulty in accurately matching the timing of the master and slave robotic arms with the stroke of the shunting machine for empty and loaded cars and the operating cycle of the tipping machine. This results in low coupler alignment accuracy and poor coordination, making it prone to abnormal conditions such as coupler tongue jamming, uneven loading, and synchronization failure. This not only reduces the overall unloading efficiency but also poses safety hazards and accelerates equipment wear, making it difficult to meet the requirements for continuous, efficient, and stable automated marshalling and unloading.

[0003] The existing technology suffers from poor coordination of coupler uncoupling operations, susceptibility to vehicle body disturbances and cycle conflicts, high jamming failure rate, and low efficiency of automated operations. Summary of the Invention

[0004] This application provides a motion coordination control method and system for automatic coupler uncoupling, which addresses the technical problems in the prior art such as poor motion coordination during coupler uncoupling operations, susceptibility to vehicle body disturbances and cycle conflicts, high jamming failure rate, and low efficiency of automated operations.

[0005] In view of the above problems, this application provides a method and system for coordinated control of automatic coupler disengagement and re-engagement.

[0006] The first aspect of this application provides a motion coordination control method for automatic coupler disengagement and re-engagement, the method comprising: Based on the coupler operating status information and the disassembly and recoupling operations, a set of motion primitives is screened and matched to generate a target motion map. The real-time position, speed, deceleration, and car body disturbance trends of the empty and loaded shunting locomotives are acquired, and car body motion constraints, tipper cycle constraints, and robotic arm collision constraints are established. Based on the target motion map and the constraints, each candidate set of motion primitives is spatiotemporally arranged to generate a spatiotemporally constrained motion set. Based on this spatiotemporally constrained motion set, a virtual coupling phase axis is constructed to perform coordinated control analysis on the motion execution process of the main and slave robotic arms, determining the cooperative lockable state intervals of the two robotic arms under the current operating cycle and coupler status. The motion execution rhythm of the main and slave robotic arms is synchronously adjusted according to the cooperative lockable state intervals, enabling the couplers at both ends to complete the uncoupling or recoupling operations according to the spatiotemporally constrained motion set.

[0007] A second aspect of this application provides a motion coordination control system for automatic coupler disengagement, the system comprising: The target motion map generation module is used to filter and match motion primitives based on coupler operating status information and disassembly actions of uncoupling and recoupling operations to generate a target motion map. The spatiotemporal constraint motion set generation module is used to acquire the real-time position, speed, deceleration, and car body disturbance trends of empty and loaded shunting locomotives, establish car body motion constraints, tipper cycle constraints, and robotic arm collision constraints, and spatiotemporally arrange each candidate motion primitive set based on the target motion map and constraints to generate a spatiotemporal constraint motion set. The state interval determination module is used to construct a virtual coupling phase axis based on the spatiotemporal constraint motion set, perform coordinated control analysis on the motion execution process of the master robotic arm and the slave robotic arm, and determine the cooperative lockable state interval of the two robotic arms under the current operating cycle and coupler status. The synchronization adjustment module is used to synchronize the motion execution rhythm of the master robotic arm and the slave robotic arm according to the cooperative lockable state interval, so that the couplers at both ends complete the uncoupling or recoupling operation according to the spatiotemporal constraint motion set.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: Based on the coupler operating status information and the disassembly and recoupling operations, a set of motion primitives is selected and matched to generate a target motion map. Real-time position, speed, deceleration, and car body disturbance trends of the empty and loaded shunting locomotives are acquired, and car body motion constraints, tipper cycle constraints, and robotic arm collision constraints are established to generate a spatiotemporal constraint motion set. A virtual coupling phase axis is constructed to determine the cooperative lockable state range of the two robotic arms under the current operating cycle and coupler status. The execution rhythm of the main and slave robotic arms is synchronously adjusted so that the couplers at both ends complete the uncoupling or recoupling operations according to the aforementioned spatiotemporal constraint motion set. This achieves precise coordination between the main and slave robotic arms and the car body movement and tipper cycle, improving the reliability and efficiency of automatic coupler uncoupling and recoupling operations. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of the motion coordination control method for automatic coupler disengagement provided in an embodiment of this application; Figure 2 This is a schematic diagram of the action coordination control system for automatic coupler disengagement and reassembly provided in an embodiment of this application.

[0011] Figure labeling: Target motion map generation module 10, spatiotemporal constraint motion set generation module 20, state interval determination module 30, synchronization adjustment module 40. Detailed Implementation

[0012] This application provides a method and system for coordinated action control of automatic coupler uncoupling, which addresses the technical problems in the prior art such as poor coordination of coupler uncoupling operations, susceptibility to vehicle body disturbances and cycle conflicts, high jamming failure rate, and low efficiency of automated operations.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] Example 1, as Figure 1 As shown, this application provides a motion coordination control method for automatic coupler disengagement, the method comprising: Step S100: Based on the coupler working condition information and the disassembly and reassembly operations, perform motion element set screening and matching to generate a target motion map.

[0015] Specifically, the system collects vehicle structure parameters, coupler opening and closing status, boom free travel, vehicle body relative posture, shunting machine speed trend, and robotic arm end contact force information to construct a coupler coupling state vector characterizing the current uncoupling and recoupling operation conditions. This vector serves as the coupler operating condition information. Simultaneously, the entire coupler uncoupling and recoupling operation process is broken down into indivisible basic execution units, forming action primitives. Each action primitive is configured with entry, hold, exit, and prohibition conditions, establishing a standardized action primitive library. Subsequently, based on the coupler coupling state vector, a set of candidate action primitives that satisfy the current state transition conditions is selected and matched from the action primitive library. Finally, a directed graph structure target action graph is generated, consisting of execution sequence relationships, state transition relationships, and corresponding trigger conditions, providing objects to be planned for subsequent spatiotemporal constraint-based action orchestration.

[0016] Step S200: Obtain the real-time position, speed, deceleration and car body disturbance trend of empty and loaded shunting locomotives, establish car body motion constraints, tipper cycle constraints and robotic arm collision constraints, and perform spatiotemporal arrangement of each candidate action primitive set based on the target motion map and the constraint conditions to generate a spatiotemporal constrained action set.

[0017] Specifically, real-time position, speed, deceleration, and car body disturbance trend information of empty and loaded shunting locomotives are collected. Based on this, the arrival time, stable period, and range of change of the working surface position of the target coupler within the planning time domain are predicted. Combined with the tippler's operating cycle, the current pose of the robotic arm, and the spatial boundary of the working area, a spatiotemporal constraint model is constructed, which includes car body motion constraints, tippler cycle constraints, robotic arm reachability constraints, and robotic arm collision constraints. Then, the entry conditions, exit conditions, action duration, and connection relationships of each candidate action element in the target motion map are read. The spatiotemporal constraint model is used to solve the executable time window, reachable spatial region, and spatial transition path of each action element, forming a spatiotemporally feasible solution set. After time conflict detection, spatial interference detection, and state transition consistency verification, infeasible combinations are eliminated, and legal combinations are arranged in an orderly manner. Finally, a spatiotemporally constrained action set containing the execution start and end times, end target pose, spatial transition trajectory, and action switching trigger conditions is generated.

[0018] Step S300: Based on the spatiotemporal constraint action set, construct a virtual coupled phase axis, perform collaborative control analysis on the action execution process of the master robot arm and the slave robot arm, and determine the collaborative lockable state range of the two robots arm in the current work cycle and coupler state.

[0019] Specifically, the execution sequence, start and end times, duration, target pose, and action switching conditions of the corresponding action primitives of the master and slave robotic arms in the spatiotemporal constraint action set are read. Key action nodes that can characterize the changes in the coupler state are extracted from these. A virtual coupled phase axis is constructed based on the time and state transition relationship of the key action nodes to uniformly characterize the opening and closing evolution process of the near-end coupler and the far-end coupler. During the execution of the dual robotic arms, the current action primitives, execution progress, actual end pose, and corresponding coupler state of the master and slave robotic arms are collected in real time. The above information is mapped to the virtual coupled phase axis to obtain the phase values ​​of the master and slave robotic arms. Based on the phase values ​​of the master and slave robotic arms, the remaining time of the current work cycle, and the opening and closing states of the couplers at both ends, a cooperative constraint domain is constructed, which is jointly determined by the phase range, the coupler state range, and the relative motion range of the vehicle body. When the phase combination of the dual robotic arms meets the preset locking conditions and the state falls within the cooperative constraint domain, the current time period is determined as the cooperative lockable state interval.

[0020] Step S400: Synchronously adjust the action execution rhythm of the master robot arm and the slave robot arm according to the cooperative lockable state interval, so that the couplers at both ends complete the uncoupling or recoupling operation according to the spatiotemporal constraint action set.

[0021] Specifically, the tippler provides the working rhythm and cycle for unloading the whole car. The empty car shunting machine is responsible for traction of empty cars and the loaded car shunting machine is responsible for traction of loaded cars. The two work together to drive the car body into place and form the relative position and motion disturbance of the car body. The main robotic arm controls the coupler of the car body at the near end, and the secondary robotic arm controls the coupler of the car body at the far end. The couplers at both ends are locked or unlocked by components such as hook tongue, locking pin, and lifting rod. The system reads the execution time window, target pose, and action switching conditions corresponding to each action primitive in the spatiotemporal constraint action set. It obtains the current phase value and corresponding coupler state of the master and slave robotic arms, matches the current phase value with the cooperative lockable state interval, determines the phase relationship between the two robotic arms, and identifies the rhythm adjustment target. The master and slave robotic arms adopt a master-slave cooperative cooperation mode. The master robotic arm approaches, aligns, and lifts the coupler lever first according to the planned action, while the slave robotic arm synchronously follows, aligns, and pushes the coupler. The actions at both ends maintain strict correspondence in timing and pose. Within the time window constraints corresponding to each action primitive, the system adjusts the execution speed, hold duration, or action of the master and / or slave robotic arms. The switching time is synchronized to reduce or eliminate phase deviation and maintain the action switching sequence. During the adjustment process, the phase difference change trend is continuously monitored to make the execution process of the two robotic arms converge towards the cooperative lockable state range. During re-coupling, the master and slave robotic arms drive the couplers at both ends to approach each other, align the hook tongues for engagement, and then push the locking pin to fall to complete the locking. During uncoupling, the master and slave robotic arms simultaneously pull the coupler lifting rod, drive the locking pin to move up to unlock the hook tongue, and then drive the coupler to separate. When the two robotic arms enter the cooperative lockable state range, a stable phase relationship is maintained and subsequent locking or uncoupling actions are executed, so that the couplers at both ends can reliably complete the uncoupling or re-coupling operation in strict accordance with the spatiotemporal constraint action set.

[0022] In one possible implementation, step S100 further includes: Step S110: Collect vehicle structure parameters, coupler opening and closing status, lifting rod free travel, vehicle body relative posture, shunting machine speed trend, and robotic arm end contact force information to construct a coupler coupling state vector representing the current uncoupling and recoupling conditions, which serves as the coupler working condition state information.

[0023] Step S120: Decompose the removal and recovery operation into indivisible action primitives, associate each action primitive with entry conditions, hold conditions, exit conditions, and prohibition conditions, and establish an action primitive library.

[0024] Step S130: Based on the coupler coupling state vector, select a set of candidate action primitives in the action primitive library that meet the current state transition conditions, and generate a target action map that meets the current state constraints.

[0025] Specifically, by acquiring data in real time from multiple sensors and reading system configurations, the system obtains vehicle structure parameters, coupler opening and closing status, boom free travel, vehicle body relative posture, shunting machine speed trend, and robotic arm end contact force information. This multi-dimensional state information is then uniformly quantified and fused to construct a coupler coupling state vector that can comprehensively and accurately characterize the current uncoupling or recoupling operation condition. This coupler coupling state vector is then used as the coupler operating condition state information for subsequent action element screening and matching.

[0026] The complete process of coupler uncoupling and recoupling is broken down into the smallest indivisible action units according to mechanical execution logic, including basic action units such as alignment and approach, lifting and unlocking, locking and clamping, retraction and avoidance, and attitude calibration. For each action unit, entry conditions, holding conditions, exit conditions, and prohibition conditions are set to constrain the start time, duration requirements, termination rules, and mutually exclusive scenarios of the action. All action units with constraints are uniformly classified and stored to form a dynamic action unit library that can be called and adapted to different working conditions.

[0027] By employing state vector matching and directed graph traversal, the real-time output coupler coupling state vector is compared item by item with the condition thresholds of each action element in the action element library. Available action units that meet the current state transition conditions are selected to form a candidate action element set. Then, based on the operation time sequence dependency and state jump logic, directed path planning and legality verification are performed on the candidate action elements to automatically generate a target action graph that meets the current working condition constraints and includes the action sequence and transition relationship.

[0028] In one possible implementation, step S100 further includes: The target action graph is a directed state graph structure, including: the execution order relationship, state transition relationship, and corresponding triggering and exit conditions of each candidate action primitive. It is used to provide objects to be planned for subsequent action orchestration under spatiotemporal constraints. Each action primitive corresponds to at least two optional successor action primitives. The selection of successor action primitives is determined by the coupler coupling state vector and the anomaly identification result.

[0029] Specifically, the target motion graph adopts a directed state graph structure. This graph clearly records the execution order relationships between candidate motion primitives, the state transition relationships between actions, and the trigger and exit conditions for each motion primitive. Its purpose is to provide directly usable motion objects for subsequent motion sequence arrangement and trajectory planning under spatiotemporal constraints. This target motion graph has branch selection capabilities, with each motion primitive corresponding to at least two selectable successor motion primitives. During operation, the specific successor motion primitive selected for jump execution is determined jointly by the coupler coupling state vector reflecting the coupler's working condition in real time and the system's online anomaly identification results. Among them, the anomaly identification result refers to the anomaly judgment information obtained by the system during operation based on real-time sensor data such as end torque of the robotic arm, joint current, displacement response, coupler contact force and vehicle body disturbance, after feature calculation and working condition judgment. This includes whether there is an anomaly, the type of anomaly, such as rust static friction lock, vertical jamming, lateral misalignment, load pre-tightening jamming, vehicle body disturbance exceeding limits, etc., the degree of anomaly and the location of the anomaly. It is used to determine whether it is necessary to jump to anomaly handling action units such as unlocking, micro-shaking, and retraction, so as to ensure that the action path can be dynamically and adaptively adjusted according to the on-site working conditions.

[0030] In one possible implementation, step S200 further includes: Step S210: Based on the real-time position, speed, deceleration and vehicle body disturbance trend information, predict the arrival time, stable period and working face position change range of the target coupler within the preset planning time domain.

[0031] Step S220: Based on the prediction results, combined with the tipper operation cycle information, the current pose information of the robotic arm, and the spatial boundary information of the coupler operation area, a spatiotemporal constraint model is constructed, including vehicle motion constraints, cycle time constraints, robotic arm accessibility constraints, and collision avoidance constraints.

[0032] Step S230: Obtain the entry conditions, exit conditions, action duration and connection relationship of each candidate action primitive in the target action graph. Combined with the spatiotemporal constraint model, solve the executable time window, reachable spatial region and spatial transition path between each candidate action primitive and adjacent action primitives to obtain a spatiotemporally feasible solution set.

[0033] Step S240: Based on the action sequence dependency, perform time conflict detection, spatial interference detection, and state transition consistency verification on the spatiotemporal feasible solution set, eliminate inoperable action combinations, and sort and arrange the remaining action combinations to generate a spatiotemporal constrained action set. The spatiotemporal constrained action set includes the execution start time, execution end time, end target pose, spatial transition trajectory, and action switching triggering conditions of each action primitive.

[0034] Specifically, based on real-time position, speed, deceleration, and car body disturbance trend information collected from empty and loaded shunting locomotives, a Long Short-Term Memory (LSTM) time-series prediction model is used to predict the target coupler state. This model takes multi-dimensional time-series sensor sequences as input and includes an input layer, a double-layer LSTM hidden layer, a fully connected layer, and an output layer. A training set is pre-constructed using on-site collected shunting locomotive operation data, car body disturbance data, and coupler positioning data. The network is trained using supervised learning, and weights are optimized through backpropagation, enabling the model to accurately fit the nonlinear motion and disturbance characteristics of the car body. The model performs rolling predictions of the coupler motion trajectory within a pre-planned time domain, determining the positioning time based on the position convergence point, identifying stable periods based on continuous intervals where the disturbance amplitude is below a set threshold, and calculating the envelope based on the upper and lower boundaries of the predicted trajectory to output the positional change range of the coupler working surface within the planned time domain. This provides high-precision time and space prediction results for subsequent spatiotemporal constraint modeling.

[0035] Based on the obtained coupler arrival time, stable period, and range of change in the working surface position, the model simultaneously incorporates the tippler operation cycle information, the current posture information of the robotic arm, and the spatial boundary information of the coupler operation area as model inputs. The constraints in the operation process are uniformly modeled into a four-dimensional constraint system to construct a spatiotemporal constraint model. Among them, the car body motion constraint is used to limit the displacement, velocity, and acceleration boundaries of the shunting locomotive and the car body within the allowable motion range; the cycle time constraint is used to strictly align the coupler uncoupling operation with the tippler operation cycle on the time axis; the robotic arm accessibility constraint is used to limit the effective working space range of the robotic arm joint movement and the end effector; and the collision avoidance constraint is used to set the minimum safe distance and space prohibition zone between robotic arms, between the robotic arm and the car body and the coupler, so that all motion primitives are executed under the premise of safety, compliance, and matching the on-site cycle.

[0036] The entry conditions, exit conditions, preset action duration, and sequential connections and dependencies between actions are extracted from the target motion graph for each candidate action primitive. These action logic parameters are then input into the constructed spatiotemporal constraint model. The model is solved using an A* path search combined with a temporal constraint planning algorithm. The A* algorithm starts with the current pose of the robotic arm and ends with the target pose of the action primitive, using spatial distance and safety cost as heuristic functions. It traverses and searches for reachable spatial regions and smooth spatial transition paths that satisfy the robotic arm's reachability constraints and collision avoidance constraints. The temporal constraint planning calculates the executable time window for each action primitive that satisfies all state transition conditions by performing temporal pruning and interval matching on vehicle motion constraints and cycle time constraints. The spatial path solution and temporal window calculation are performed on all candidate action primitives. The legal solutions that simultaneously satisfy the temporal constraints, spatial constraints, and action connections are summarized, ultimately forming a spatiotemporally feasible solution set consisting of multiple sets of feasible time windows, reachable spatial regions, and transition paths.

[0037] Based on the sequential dependencies of actions in the target motion graph, multi-dimensional compliance checks are performed on each set of action schemes within the spatiotemporal feasible solution set: time conflict detection is completed by determining temporal overlap to determine whether there are problems such as reversed or overlapping execution time windows of each action primitive that violate the work rhythm and vehicle motion constraints; spatial interference detection is performed by spatial bounding box intersection and distance discrimination method to verify whether there is collision intrusion between the robotic arm end effector, transition trajectory and vehicle body, coupler and work area boundaries; at the same time, state transition consistency verification is carried out by combining the coupler coupling state vector and the condition rules of action primitives to confirm whether the current action exit condition and the next action entry condition match and whether the state jump conforms to the working condition logic. Action combinations that have time conflicts, spatial interference, or discontinuous state transitions during detection and verification are directly eliminated. The remaining compliant and feasible action combinations are sorted, optimized, and arranged in an overall manner according to the priority of the operation sequence and the spatial smoothness index. Finally, a standardized spatiotemporal constraint action set is generated. This action set fully includes the execution start time, execution end time, target pose of the robotic arm end, continuous and smooth spatial transition trajectory, and action switching trigger conditions of each action primitive, which serve as the direct execution basis for the master and slave robotic arms to coordinate the unhooking and re-hooking operation.

[0038] In one possible implementation, step S300 further includes: Step S310: Read the execution order, start and end times, duration, target pose, and action switching conditions of each action primitive corresponding to the master and slave robotic arms in the spatiotemporal constraint action set.

[0039] Step S320: Extract key motion nodes representing the changes in coupler state from the motion primitives, and construct a virtual coupled phase axis based on the time relationship and state transition relationship of the key motion nodes to represent the opening and closing evolution process of the near-end coupler and the far-end coupler.

[0040] Step S330: During the execution of the dual robotic arms, the current motion primitives, motion execution progress, actual end pose and corresponding coupler status of the master robotic arm and slave robotic arm are collected in real time. The collected results are mapped to the virtual coupling phase axis to obtain the phase value of the master robotic arm and the phase value of the slave robotic arm.

[0041] Step S340: Based on the phase value of the main robotic arm, the phase value of the slave robotic arm, the remaining time of the current work cycle, and the opening and closing status of the couplers at both ends, determine the cooperative execution relationship of the two robotic arms. When it is determined that the phase combination of the main robotic arm and the slave robotic arm meets the preset locking condition, the current time period is determined as the cooperative locking state interval.

[0042] Specifically, from the generated spatiotemporal constraint action set, the instruction information of each action primitive assigned to the master robot arm and the slave robot arm is parsed and read, including the execution order of each action primitive in collaborative operation, the pre-planned execution start time and execution end time, the duration required for the corresponding action, the target pose to be reached by the robot arm end effector, and the action switching conditions that trigger the action primitive jump. The above data is uniformly loaded into the collaborative controller to provide standardized timing and pose reference data for subsequent dual robot arm phase coupling and collaborative execution.

[0043] Key motion nodes that can directly characterize the state changes of the coupler lifting arm, locking tongue rotation, coupler engagement and disengagement are selected and extracted from all motion primitives. Based on the chronological order of these key motion nodes in the operation process and the state transition logic corresponding to coupler opening and closing, a unified virtual coupling phase axis is constructed. This phase axis uses phase scale to quantitatively characterize the complete evolution process of the near-end coupler and the far-end coupler from alignment, approach, unlocking to locking, so that the real-time operation status of the couplers at both ends can be expressed in phase and compared in the same dimension, providing a unified reference benchmark for the synchronous monitoring of the subsequent dual robotic arm execution progress.

[0044] During the collaborative execution of coupler uncoupling and rejoining operations by the main and slave robotic arms, the joint encoder, visual positioning sensor, and coupler status detection unit collect in real time the current action element type, percentage of action progress, actual pose data of the end effector, and the real-time opening and closing status of their respective near and far couplers. The above multi-dimensional real-time data are matched and projected onto the constructed virtual coupled phase axis according to a preset mapping relationship. The phase value of the main robotic arm and the current phase value of the slave robotic arm are obtained by phase interpolation calculation, realizing a unified phase-quantified expression of the dual robotic arm operation progress and coupler status.

[0045] The system takes four parameters as input: the phase value of the master robotic arm, the phase value of the slave robotic arm, the remaining time of the current work cycle, and the opening and closing status of the near and far couplers. It conducts a comprehensive judgment by constructing a collaborative judgment logic table and a phase combination threshold range. First, it calculates the phase difference and phase synchronization of the master and slave robotic arms, verifies whether the execution progress of the motion primitives at both ends matches, and determines whether the remaining time of the work cycle meets the minimum duration requirement for the locking action. It also verifies whether the couplers at both ends are in a lockable physical posture. When all the above conditions are met and the phase combination of the master and slave robotic arms falls within the pre-set qualified locking phase range, it is determined that the preset locking conditions are met. The current continuous stable period is marked as the collaborative lockable state range, providing accurate timing judgment results for the dual robotic arms to perform synchronous locking operations.

[0046] In one possible implementation, step S400 further includes: Step S410: Read the execution time window, target pose, and action switching conditions of each action primitive in the spatiotemporal constraint action set.

[0047] Step S420: Obtain the current phase value and corresponding coupler status of the master robot arm and slave robot arm, and match them based on the current phase value and the cooperative lockable state interval to determine the phase relationship between the two robot arms and the corresponding rhythm adjustment target.

[0048] Step S430: Within the time window constraints corresponding to each motion primitive, adjust the execution speed, hold duration, or motion switching time of the master robot arm and / or slave robot arm to reduce or eliminate phase deviation and maintain the motion switching sequence unchanged.

[0049] Step S440: During the adjustment process, continuously monitor the phase difference change trend to make the execution process of the dual robotic arms converge towards the cooperative lockable state range.

[0050] Step S450: When the two robotic arms enter the cooperative lockable state range, maintain their phase relationship and perform subsequent locking or unhooking actions.

[0051] Specifically, from the generated and verified spatiotemporal constraint action set, the planning parameters corresponding to each action primitive assigned to the master and slave robotic arms are parsed and read one by one. These parameters include the start-to-end time window for each action primitive to be executed, the spatial target pose that the robotic arm end effector needs to reach, including position coordinates and attitude angles, and the action switching conditions that can trigger the action jump when the state is met. The above parameters are then loaded into the cooperative motion controller as the reference constraints and instruction basis for subsequent dual-arm phase synchronization and action execution.

[0052] The system acquires the current phase values ​​of the master and slave robotic arms on the virtual coupling phase axis in real time, and simultaneously collects the real-time opening and closing status of the near-end couplers and far-end couplers corresponding to each robotic arm. The two sets of phase values ​​are substituted into a preset cooperative lockable interval for interval determination, and the difference between the phase values ​​of the master and slave robotic arms is calculated. When the phase difference is within a set small threshold range, the two robotic arms are determined to be in a synchronized state. When the phase value of the master robotic arm is greater than that of the slave robotic arm and the difference exceeds the synchronization threshold, the master robotic arm is determined to be ahead and the slave robotic arm is behind. When the phase value of the slave robotic arm is greater than that of the master robotic arm and the difference exceeds the synchronization threshold, the slave robotic arm is determined to be ahead and the master robotic arm is behind. Based on the above determination results, corresponding rhythm adjustment targets are generated, such as the master robotic arm decelerating and maintaining, and the slave robotic arm accelerating to catch up, or the slave robotic arm appropriately slowing down and the master robotic arm slightly accelerating, providing a clear control direction for subsequent phase synchronization correction.

[0053] Within the pre-planned execution time window constraints of each motion primitive, a phase closed-loop proportional adjustment strategy is adopted. Based on the direction and magnitude of the phase deviation determined in the previous step, the joint movement speed, intermediate posture holding time, or motion switching trigger time of the master robot arm, slave robot arm, or both are dynamically adjusted. When the phase of the master robot arm is ahead, its running speed is appropriately reduced and its posture holding time is extended. When the phase of the slave robot arm is behind, its execution speed is moderately increased. Under the premise of strictly ensuring that the original motion switching sequence is not disrupted and does not exceed the allowable time window boundary, the phase deviation between the two arms is gradually reduced and eliminated, so that the execution progress of the two tends to be consistent.

[0054] During the dynamic phase adjustment of the dual-arm system, real-time phase difference closed-loop monitoring and gradient convergence control are adopted. The phase values ​​of the master and slave robotic arms are continuously sampled and the real-time phase difference is calculated. By performing a moving average on the phase difference data of multiple consecutive cycles, the shrinking or diverging trend of the phase deviation is identified. The adjustment intensity is adaptively fine-tuned according to the deviation convergence rate to avoid overshoot and oscillation. At the same time, the adjustment process is constrained to never exceed the action time window and spatial constraints, gradually and smoothly pushing the execution progress of the dual robotic arms into the cooperative lockable state range and achieving progressive convergence.

[0055] Once the phase combination of the master and slave robotic arms has been stably placed within the cooperative lockable state range by real-time phase monitoring, the phase synchronization maintenance logic is immediately activated to strictly lock the phase difference and relative timing relationship between the two arms and maintain consistent action rhythm. Subsequently, according to the predetermined process of the spatiotemporal constraint action set, the locking or uncoupling action of the couplers at both ends is triggered and executed synchronously to ensure that the two robotic arms complete the final linkage operation under the premise of phase coordination, thus ensuring operational safety and synchronization accuracy.

[0056] In one possible implementation, step S430 further includes: Step S431: The action unit includes a jittering vibration unit, wherein, before the lifting and unlocking unit is executed, a trial vibration with an amplitude smaller than that of the jittering vibration unit is applied by the end of the robotic arm, and the corresponding force or displacement response data is collected.

[0057] Step S432: Evaluate the equivalent stiffness and equivalent damping of the coupler contact pair based on the force or displacement response data, and determine the type of jamming mechanism in the current contact state.

[0058] Step S433 classifies the shaking and swaying operation according to the type of jamming mechanism. When it is determined to be static friction lock, the shaking and swaying basic element is called; when it is determined to be off-center load jamming, the lateral micro-shaking basic element or the longitudinal retraction basic element is called.

[0059] Specifically, the motion unit includes a shaking and oscillating unit specifically designed to release coupler jamming. Before entering the formal lifting and unlocking unit execution stage, the end effector of the robotic arm is controlled to apply a set of low-amplitude, small-amplitude probing vibrations to the coupler. The amplitude of these probing vibrations is strictly less than the vibration amplitude of the shaking and oscillating unit during normal operation to avoid impacting the coupler structure. At the same time, the six-dimensional force sensor and displacement detection unit integrated at the end of the robotic arm collect the force feedback response data or displacement response data of the coupler contact pair under the action of the probing vibrations in real time, providing the original detection basis for subsequent jamming state judgment.

[0060] The collected force and displacement time-series data under experimental vibration were substituted into a single-degree-of-freedom mass-spring-damped dynamic model. With displacement as the excitation input and force signal as the output response, the system transfer function was solved by fitting using the least squares method, directly identifying the equivalent stiffness and equivalent damping parameters of the coupler contact pair. Based on this, the jamming mechanism was determined according to the parameter characteristics: if the equivalent stiffness is large and the equivalent damping is stable, it is determined to be static friction locking; if the stiffness shows a lateral asymmetric distribution and the displacement response has obvious lag, it is determined to be eccentric loading jamming; if the stiffness shows a step change and the force signal is accompanied by jumps and burrs, it is determined to be structural interference jamming, thus completing the accurate differentiation of the jamming type in the contact state.

[0061] Based on the identified jamming mechanism type, differentiated and categorized shaking and swaying unblocking operations are performed: if it is determined to be static friction lock of the contact surface, the preset shaking and swaying element is directly called to break the static friction through high-frequency small-amplitude reciprocating vibration, so that the coupler contact pair enters a sliding state that is easy to move; if it is determined to be jamming due to uneven lateral force, the lateral micro-shaking element is called to perform lateral attitude correction according to the direction of the uneven load and the degree of jamming, or the longitudinal retraction element is called to first slightly retract to release the contact pressure and eliminate squeezing interference, and then cooperate with micro-motion adjustment to achieve unblocking, clearing obstacles for the smooth execution of the subsequent lifting and unlocking element.

[0062] In one possible implementation, step S430 further includes: During the execution of the spatiotemporal constraint action set, the end torque, joint current and displacement response data of the robotic arm are collected in real time.

[0063] Based on the collected data on end-effector torque, joint current, and displacement response of the robotic arm, the contact stiffness characteristics, resistance torque characteristics, and displacement increment characteristics are calculated.

[0064] When the contact stiffness characteristics, resistance torque characteristics, and displacement increment characteristics meet the preset abnormal working condition criteria, the current abnormality is identified as belonging to at least one of the following: rust static friction lock-up, vertical jamming, lateral seizure, or load pre-tightening jamming.

[0065] Based on the exception type, the corresponding micro-adjustment primitive combination is invoked to reconstruct the current action sequence online. After the exception is resolved, the original state transition path is returned to continue the unhooking or rehooking operation.

[0066] Specifically, during the entire execution of the spatiotemporal constraint action set, the six-dimensional force sensor integrated at the end of the robotic arm collects the end torque data in real time, the servo driver collects the current feedback signals of each joint, and the joint encoder continuously collects the displacement response data of the end contacting the coupler. The three types of signals are synchronously sampled and filtered to form a real-time sensing data stream for working condition monitoring.

[0067] Based on the collected time-series data of end-effector torque, joint current, and displacement response of the robotic arm, the following steps are taken: First, the real-time torque of the robotic arm end-effector acting on the coupler is used as input, and the real-time displacement generated by the coupler contact pair is used as output. The ratio of the end-effector torque and the corresponding displacement at the same moment is calculated to obtain the contact stiffness characteristics, which reflect the rigidity and clamping degree of the coupler contact pair. The actual output torque of the robotic arm acting on the coupler is obtained by converting the joint current with the torque constant of the servo motor and the transmission ratio of the robotic arm. Combined with the change in the rotation angle or linear displacement of the coupler, the resistance torque characteristics are extracted to characterize the magnitude of the resistance encountered by the coupler's rotation or movement. At the same time, a first-order difference operation is performed on the displacement data sampled at continuous equal time intervals, that is, the displacement at the next moment is subtracted from the displacement at the previous moment, and then divided by the sampling time interval to obtain the displacement increment characteristics per unit time. This is used to determine whether the coupler has actual movement or is completely jammed, providing a quantitative basis for subsequent abnormal working condition identification.

[0068] A multi-feature threshold joint judgment and working condition mode matching method is adopted to compare the real-time calculated contact stiffness, resistance torque, and displacement increment with a preset threshold library: if the contact stiffness is greater than the preset high stiffness threshold, the resistance torque is greater than the preset stall threshold, and the displacement increment is less than the preset fretting threshold, and all three conditions are met simultaneously, it is judged as corrosion static friction lock-up; if the vertical contact stiffness is greater than the preset vertical stiffness threshold, the vertical displacement increment is approximately zero and the duration exceeds the preset stagnation time threshold, it is judged as vertical jamming; if the difference between the left and right lateral torques is greater than the preset eccentric load torque threshold, the lateral displacement response lag time exceeds the preset lag threshold, and the displacement deviation is greater than the preset deviation threshold, it is judged as lateral misalignment; if the contact stiffness increases linearly with the increase of external force, the resistance torque increases synchronously, and the displacement increment is continuously less than the preset low increment threshold, it is judged as preloaded jamming, thereby achieving accurate identification of abnormal types.

[0069] Based on the identified anomaly type, a dynamic scheduling strategy combining primitives is adopted: for rust static friction locking, high-frequency jittering and shaking primitives are called; for vertical jamming, longitudinal micro-lifting and retraction primitives are called; for lateral misalignment, left and right micro-swaying primitives are called; and for pre-tightening jamming under load, progressive loosening primitives are called. In this way, the current action sequence is reconstructed online. During the execution process, characteristic values ​​are continuously monitored. When the contact stiffness, resistance torque, and displacement increment all return to the normal threshold range and the anomaly criteria are not met, the adjustment process is immediately terminated, and the state transition path of the original spatiotemporal constraint action set is switched back to continue the unhooking or rehooking operation according to the original plan.

[0070] In one possible implementation, step S340 further includes: Step S341: Construct a cooperative constraint domain characterizing the cooperative state of the couplers at both ends. The cooperative constraint domain is jointly determined by the phase range of the master robot arm and the slave robot arm, the opening and closing state range of the couplers at both ends, and the relative motion state range of the vehicle body.

[0071] Step S342: Map the current phase values ​​of the master robot arm and the slave robot arm, the corresponding coupler state, and the current relative motion state of the vehicle body to the cooperative constraint domain.

[0072] Step S343: When the current states of the master robotic arm and the slave robotic arm both fall within the cooperative constraint domain, determine that the current time period is the cooperative lockable state interval or the cooperative unhooking allowed interval, and allow the execution of the corresponding key action according to the spatiotemporal constraint action set.

[0073] Specifically, a collaborative constraint domain is constructed to quantitatively characterize the coordinated operation state of the couplers at both ends controlled by the master and slave robotic arms. The collaborative constraint domain is a three-dimensional permissible state space that satisfies the requirements for safe coupler engagement and reliable coupler disengagement. The phase range between the master and slave robotic arms refers to the permissible synchronization deviation range of the end effector's motion timing and progress during coupler alignment, lifting, and pushing actions, ensuring the coordination of the two arms' movements. The coupler opening / closing state range refers to the operable state range corresponding to the coupler tongue rotation angle, locking pin position, and coupler cavity engagement depth, within which the coupler possesses the structural conditions to perform engagement or disengagement. The relative motion state range of the car bodies refers to the safe permissible range of relative displacement, relative velocity, and relative acceleration between adjacent car bodies, preventing alignment failure or impact damage due to excessive relative motion. The state ranges of these three dimensions together constitute the boundary conditions of the collaborative constraint domain, serving as the basis for subsequent collaborative operation permission determination.

[0074] The real-time collected phase values ​​of the master robotic arm, slave robotic arms, coupler states corresponding to each robotic arm, and relative motion states of adjacent vehicles are mapped to the aforementioned collaborative constraint domain through coordinate normalization and dimensional unification. The current phase value represents the real-time progress parameters of the master and slave robotic arms within their respective action cycles, characterizing the temporal position of the two arms' movements. The corresponding coupler state is a set of hook tongue angles, locking pin positions, engagement states, and force states of the couplers at both ends, reflecting whether the couplers meet the operational conditions. The relative motion states of the vehicles are the real-time relative displacement, relative velocity, and relative impact between adjacent vehicles, characterizing the vehicle coupling motion characteristics. By mapping these three types of state variables to the phase dimension, coupler state dimension, and vehicle motion dimension of the collaborative constraint domain, the current collaborative operation state is located within the constraint domain.

[0075] Real-time boundary determination is performed on the current states of the master and slave robotic arms mapped to the collaborative constraint domain. The current state includes a real-time combination of the phase values ​​of the master and slave robotic arms, the opening and closing states of the couplers at both ends, and the relative motion state of the vehicle body. When all the combined state points are within the three-dimensional boundary of the collaborative constraint domain and do not exceed the threshold range of any dimension, the current time period is determined to be either a collaborative locking state interval or a collaborative uncoupling allowable interval. The collaborative locking state interval refers to the allowable operation period when the couplers at both ends are accurately aligned, the movements of both arms are synchronized, and the vehicle body moves smoothly, satisfying the safety locking requirement. The collaborative uncoupling allowable interval refers to the allowable operation period when the coupler load is fully unloaded, the movement phases are consistent, and the relative displacement of the vehicle body is stable, satisfying the reliable uncoupling requirement. Within the above intervals, the system outputs operation permission instructions, allowing the actuators to strictly follow the pre-planned spatiotemporal constraint action set to complete the key collaborative actions of coupler locking or coupler uncoupling.

[0076] Example 2, based on the same inventive concept as the automatic coupler disengagement and re-engagement action coordination control method in the aforementioned examples, such as... Figure 2 As shown, this application provides a motion-coordinated control system for automatic coupler disengagement. The system and method embodiments in this application are based on the same inventive concept. The system includes: The target motion map generation module 10 is used to filter and match the motion primitive set based on the coupler working condition information and the disassembly action of the coupling operation to generate the target motion map.

[0077] The spatiotemporal constraint action set generation module 20 is used to obtain the real-time position, speed, deceleration and car body disturbance trend of empty car shunting locomotive and loaded car shunting locomotive, establish car body motion constraints, tipper cycle constraints and robotic arm collision constraints, and perform spatiotemporal arrangement of each candidate action primitive set based on the target motion map combined with the constraint conditions to generate a spatiotemporal constraint action set.

[0078] The state interval determination module 30 is used to construct a virtual coupled phase axis based on the spatiotemporal constraint action set, perform coordinated control analysis on the action execution process of the master robot arm and the slave robot arm, and determine the coordinated lockable state interval of the two robots arm in the current working cycle and coupler state.

[0079] The synchronization adjustment module 40 is used to synchronize the action execution rhythm of the master robot arm and the slave robot arm according to the cooperative lockable state interval, so that the couplers at both ends complete the uncoupling or recoupling operation according to the spatiotemporal constraint action set.

[0080] Furthermore, the system is also used to implement the following functions: Collect vehicle structure parameters, coupler opening and closing status, lifting rod free travel, vehicle body relative posture, shunting machine speed trend, and robotic arm end contact force information to construct a coupler coupling state vector representing the current uncoupling and recoupling operation, which serves as the coupler operation state information. Decompose the uncoupling and recoupling operation into indivisible action primitives, and associate each action primitive with entry conditions, holding conditions, exit conditions, and prohibition conditions to establish an action primitive library. Based on the coupler coupling state vector, select a set of candidate action primitives that meet the current state transition conditions from the action primitive library to generate a target action map that meets the current state constraints.

[0081] Furthermore, the system is also used to implement the following functions: The target action graph is a directed state graph structure, including: the execution order relationship, state transition relationship, and corresponding triggering and exit conditions of each candidate action primitive. It is used to provide objects to be planned for subsequent action orchestration under spatiotemporal constraints. Each action primitive corresponds to at least two optional successor action primitives. The selection of successor action primitives is determined by the coupler coupling state vector and the anomaly identification result.

[0082] Furthermore, the system is also used to implement the following functions: Based on the real-time position, velocity, deceleration, and vehicle body disturbance trend information, the arrival time, stable period, and range of change in the working surface position of the target coupler within the preset planning time domain are predicted. Based on the prediction results, combined with the tipper's operating cycle information, the current posture information of the robotic arm, and the spatial boundary information of the coupler's working area, a spatiotemporal constraint model is constructed, including vehicle body motion constraints, cycle time constraints, robotic arm reachability constraints, and collision avoidance constraints. The entry conditions, exit conditions, action duration, and connection relationships of each candidate action primitive in the target motion map are obtained, combined with the spatiotemporal constraints. The bundle model is used to solve for the executable time window, reachable spatial region, and spatial transition path between each candidate action primitive and adjacent action primitives, thus obtaining a spatiotemporally feasible solution set. Based on the action sequence dependency, the spatiotemporally feasible solution set is subjected to time conflict detection, spatial interference detection, and state transition consistency verification. Infeasible action combinations are eliminated, and the remaining action combinations are sorted and arranged to generate a spatiotemporally constrained action set. The spatiotemporally constrained action set includes the execution start time, execution end time, end target pose, spatial transition trajectory, and action switching trigger condition of each action primitive.

[0083] Furthermore, the system is also used to implement the following functions: The execution sequence, start and end times, duration, target pose, and action switching conditions of each action element corresponding to the master and slave robotic arms in the spatiotemporal constraint action set are read. Key action nodes representing coupler state changes are extracted from the action elements. A virtual coupled phase axis is constructed based on the temporal and state transition relationships of the key action nodes to represent the opening and closing evolution process of the near-end and far-end couplers. During the execution of the dual robotic arms, the current action elements, action execution progress, end-effector pose, and corresponding coupler states of the master and slave robotic arms are collected in real time. The collected results are mapped to the virtual coupled phase axis to obtain the phase values ​​of the master and slave robotic arms. Based on the phase values ​​of the master and slave robotic arms, the remaining time of the current work cycle, and the opening and closing states of the couplers at both ends, the cooperative execution relationship of the dual robotic arms is determined. When the phase combination of the master and slave robotic arms satisfies the preset locking conditions, the current time period is determined as a cooperative lockable state interval.

[0084] Furthermore, the system is also used to implement the following functions: The system reads the execution time window, target pose, and action switching conditions of each action primitive in the spatiotemporal constraint action set; obtains the current phase value and corresponding coupler state of the master and slave robotic arms; matches the current phase value with the cooperative lockable state interval to determine the phase relationship between the two robotic arms and the corresponding rhythm adjustment target; within the time window constraints corresponding to each action primitive, adjusts the execution speed, holding time, or action switching time of the master and / or slave robotic arms to reduce or eliminate phase deviation and maintain the action switching order unchanged; continuously monitors the phase difference change trend during the adjustment process to make the execution process of the two robotic arms converge towards the cooperative lockable state interval; when the two robotic arms enter the cooperative lockable state interval, maintain their phase relationship and execute subsequent locking or unhooking actions.

[0085] Furthermore, the system is also used to implement the following functions: The action unit includes a jittering vibration unit. Before the lifting and unlocking unit is executed, a trial vibration with an amplitude smaller than that of the jittering vibration unit is applied by the end of the robotic arm, and corresponding force or displacement response data is collected. Based on the force or displacement response data, the equivalent stiffness and equivalent damping of the coupler contact pair are evaluated, and the jamming mechanism type of the current contact state is determined. The jittering vibration operation is classified according to the jamming mechanism type. When it is determined to be static friction locking, the jittering vibration unit is invoked. When it is determined to be off-center loading jamming, the lateral micro-swaying unit or the longitudinal retraction unit is invoked.

[0086] Furthermore, the system is also used to implement the following functions: During the execution of the spatiotemporal constraint action set, real-time data on the end torque, joint current, and displacement response of the robotic arm are collected. Based on the collected end torque, joint current, and displacement response data, contact stiffness characteristics, resistance torque characteristics, and displacement increment characteristics are calculated. When the contact stiffness characteristics, resistance torque characteristics, and displacement increment characteristics meet the preset abnormal working condition criteria, the current abnormality is identified as at least one of rust static friction locking, vertical jamming, lateral misalignment, or load pre-tightening jamming. According to the abnormality type, the corresponding micro-motion adjustment element combination is called to reconstruct the current action sequence online. After the abnormality is resolved, the original state transition path is returned to continue the unhooking or rehooking operation.

[0087] Furthermore, the system is also used to implement the following functions: A collaborative constraint domain is constructed to characterize the coordinated state of the couplers at both ends. The collaborative constraint domain is jointly determined by the phase range of the master and slave manipulators, the opening and closing state range of the couplers at both ends, and the relative motion state range of the vehicle body. The current phase values ​​of the master and slave manipulators, the corresponding coupler states, and the current relative motion state of the vehicle body are mapped to the collaborative constraint domain. When the current states of the master and slave manipulators simultaneously fall within the collaborative constraint domain, the current time period is determined to be either the collaborative lockable state interval or the collaborative uncoupling allowed interval, and the corresponding key action is allowed to be executed according to the spatiotemporal constraint action set.

[0088] It should be noted that the order of the embodiments described above is for descriptive purposes only and does not represent the superiority or inferiority of the embodiments. Specific embodiments of this specification have been described above. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0089] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0090] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for coordinated control of automatic coupler disengagement, characterized in that, include: Based on the coupler working condition information and the disassembly and reassembly operations, the action element set is filtered and matched to generate the target action map. The real-time position, speed, deceleration and car body disturbance trend of empty and loaded shunting locomotives are obtained. Car body motion constraints, tipper cycle constraints and robotic arm collision constraints are established. Based on the target motion map and the constraint conditions, the set of candidate motion primitives is spatiotemporally arranged to generate a spatiotemporally constrained motion set. Based on the spatiotemporal constraint action set, a virtual coupled phase axis is constructed to perform coordinated control analysis on the action execution process of the master robot arm and the slave robot arm, and to determine the coordinated lockable state range of the two robots arm in the current working cycle and coupler state. The action execution rhythm of the master and slave robotic arms is synchronously adjusted according to the cooperative lockable state interval, so that the couplers at both ends complete the unhooking or rehooking operation according to the spatiotemporal constraint action set.

2. The action coordination control method for automatic coupler disengagement according to claim 1, characterized in that, Based on the coupler's operating condition information and the disassembly / reunification operations, a set of motion primitives is filtered and matched to generate a target motion map, including: Collect vehicle structure parameters, coupler opening and closing status, lifting rod free travel, vehicle body relative posture, shunting machine speed trend, and robotic arm end contact force information to construct a coupler coupling state vector representing the current uncoupling and recoupling working conditions, which serves as the coupler working condition state information. The removal and recovery operation is decomposed into indivisible action primitives, and each action primitive is associated with entry conditions, hold conditions, exit conditions, and prohibition conditions to establish an action primitive library; Based on the coupler coupling state vector, a set of candidate action primitives that meet the current state transition conditions are selected from the action primitive library to generate a target action map that meets the current state constraints.

3. The automatic coupler disengagement and re-engagement action coordination control method according to claim 2, characterized in that, The target action graph is a directed state graph structure, including: the execution order relationship, state transition relationship, and corresponding triggering and exit conditions of each candidate action primitive. It is used to provide objects to be planned for subsequent action orchestration under spatiotemporal constraints. Each action primitive corresponds to at least two optional successor action primitives. The selection of successor action primitives is determined by the coupler coupling state vector and the anomaly identification result.

4. The action coordination control method for automatic coupler disengagement according to claim 3, characterized in that, Based on the target action graph and constraints, the candidate action primitive sets are spatiotemporally orchestrated to generate a spatiotemporally constrained action set, including: Based on the real-time position, speed, deceleration and vehicle body disturbance trend information, the arrival time, stable period and working face position change range of the target coupler are predicted within the preset planning time domain. Based on the prediction results, combined with the tipper operation cycle information, the current pose information of the robotic arm and the spatial boundary information of the coupler operation area, a spatiotemporal constraint model is constructed, including vehicle motion constraints, cycle time constraints, robotic arm accessibility constraints and collision avoidance constraints. The entry conditions, exit conditions, action duration and connection relationships of each candidate action primitive in the target action graph are obtained. Combined with the spatiotemporal constraint model, the executable time window, reachable spatial region and spatial transition path between each candidate action primitive and adjacent action primitive are solved to obtain a spatiotemporally feasible solution set. Based on the sequential dependency of actions, the spatiotemporal feasible solution set is subjected to time conflict detection, spatial interference detection, and state transition consistency verification. Infeasible action combinations are eliminated, and the remaining action combinations are sorted and arranged to generate a spatiotemporal constrained action set. The spatiotemporal constrained action set includes the execution start time, execution end time, end target pose, spatial transition trajectory, and action switching triggering conditions of each action primitive.

5. The action coordination control method for automatic coupler disengagement according to claim 1, characterized in that, Based on the spatiotemporal constraint action set, a virtual coupled phase axis is constructed to perform coordinated control analysis on the action execution process of the master and slave robotic arms, determining the coordinated lockable state intervals of the two robotic arms under the current work cycle and coupler state, including: Read the execution order, start and end times, duration, target pose, and action switching conditions of each action primitive corresponding to the master and slave robotic arms in the spatiotemporal constraint action set; Key motion nodes representing changes in coupler state are extracted from motion primitives. Virtual coupled phase axes are constructed based on the temporal and state transition relationships of the key motion nodes to represent the opening and closing evolution process of the near-end coupler and the far-end coupler. During the execution of the dual robotic arms, the current motion primitives, motion execution progress, actual end pose and corresponding coupler status of the master robotic arm and slave robotic arm are collected in real time. The collected results are mapped to the virtual coupling phase axis to obtain the phase values ​​of the master robotic arm and slave robotic arm. Based on the phase value of the master robotic arm, the phase value of the slave robotic arm, the remaining time of the current work cycle, and the opening and closing status of the couplers at both ends, the cooperative execution relationship of the two robotic arms is determined. When it is determined that the phase combination of the master robotic arm and the slave robotic arm meets the preset locking condition, the current time period is determined as the cooperative locking state interval.

6. The action coordination control method for automatic coupler disengagement according to claim 5, characterized in that, The execution rhythm of the main and slave robotic arms is synchronously adjusted according to the cooperative lockable state interval, so that the couplers at both ends complete the uncoupling or recoupling operation according to the spatiotemporal constraint action set, including: Read the execution time window, target pose, and action switching conditions of each action primitive in the spatiotemporal constraint action set; Obtain the current phase values ​​of the master and slave robotic arms and the corresponding coupler status. Based on the current phase values ​​and the cooperative lockable state interval, perform matching to determine the phase relationship between the two robotic arms and the corresponding rhythm adjustment target. Within the time window constraints corresponding to each motion primitive, the execution speed, hold duration, or motion switching time of the master robot arm and / or slave robot arm are adjusted to reduce or eliminate phase deviation and maintain the motion switching sequence unchanged. During the adjustment process, the phase difference change trend is continuously monitored to make the execution process of the dual robotic arms converge towards the cooperative lockable state range; When the two robotic arms enter the cooperative lockable state range, they maintain their phase relationship and perform subsequent locking or unhooking actions.

7. The action coordination control method for automatic coupler disengagement according to claim 2, characterized in that, The action unit includes a shaking vibration unit, wherein, before the lifting and unlocking unit is executed, a trial vibration with an amplitude smaller than that of the shaking vibration unit is applied by the end of the robotic arm, and the corresponding force or displacement response data is collected. The equivalent stiffness and equivalent damping of the coupler contact pair are evaluated based on the force or displacement response data to determine the type of jamming mechanism in the current contact state. The shaking and swaying operations are classified according to the type of jamming mechanism. When it is determined to be static friction lock, the shaking and swaying basic element is called; when it is determined to be off-center load jamming, the lateral micro-shaking basic element or the longitudinal retraction basic element is called.

8. The action coordination control method for automatic coupler disengagement according to claim 1, characterized in that, Also includes: During the execution of the spatiotemporal constraint action set, the end torque, joint current and displacement response data of the robotic arm are collected in real time; Based on the collected data on end-effector torque, joint current, and displacement response of the robotic arm, the contact stiffness characteristics, resistance torque characteristics, and displacement increment characteristics are calculated. When the contact stiffness characteristics, resistance torque characteristics and displacement increment characteristics meet the preset abnormal working condition criteria, the current abnormality is identified as at least one of the following: corrosion static friction locking, vertical jamming, lateral seizing, or load pre-tightening jamming. Based on the exception type, the corresponding micro-adjustment primitive combination is invoked to reconstruct the current action sequence online. After the exception is resolved, the original state transition path is returned to continue the unhooking or rehooking operation.

9. The action coordination control method for automatic coupler disengagement according to claim 5, characterized in that, Determining the cooperative execution relationship between the two robotic arms includes: A collaborative constraint domain is constructed to characterize the collaborative state of the couplers at both ends. The collaborative constraint domain is jointly determined by the phase range of the master robot arm and the slave robot arm, the opening and closing state range of the couplers at both ends, and the relative motion state range of the vehicle body. Map the current phase values ​​of the master and slave robotic arms, the corresponding coupler states, and the current relative motion state of the vehicle body to the cooperative constraint domain; When the current states of the master robotic arm and the slave robotic arm both fall within the cooperative constraint domain, the current time period is determined to be either the cooperative lockable state interval or the cooperative unhooking allowed interval, and the corresponding key action is allowed to be executed according to the spatiotemporal constraint action set.

10. An automatic coupler disengagement and re-engagement action coordination control system, characterized in that, The system is used to implement the motion coordination control method for automatic coupler disengagement as described in any one of claims 1-9, the system comprising: The target motion map generation module is used to filter and match the motion primitive set based on the coupler working condition information and the disassembly action of the coupling operation to generate the target motion map. The spatiotemporal constraint action set generation module is used to obtain the real-time position, speed, deceleration and car body disturbance trend of empty car shunting locomotive and loaded car shunting locomotive, establish car body motion constraints, tipper cycle constraints and robotic arm collision constraints, and perform spatiotemporal arrangement of each candidate action primitive set based on the target motion map and the constraint conditions to generate a spatiotemporal constraint action set. The state interval determination module is used to construct a virtual coupled phase axis based on the spatiotemporal constraint action set, perform coordinated control analysis on the action execution process of the master robot arm and the slave robot arm, and determine the coordinated lockable state interval of the two robots arm in the current working cycle and coupler state. The synchronization adjustment module is used to synchronize the action execution rhythm of the master robotic arm and the slave robotic arm according to the cooperative lockable state interval, so that the couplers at both ends can complete the unhooking or rehooking operation according to the spatiotemporal constraint action set.

Images