Cooperative control method and system for dumper robots used in train operation lines

Abstract

This application provides a collaborative control method and system for tipper robots on train operation lines, relating to the field of robot control technology. It involves collecting a set of task features for the tipper operation; using these features as constraints, analyzing the task allocation of the tipper robots to obtain collaborative task planning, including unhooking planning for the unhooking robot, hooking planning for the hooking robot, and hooking planning for the re-hooking robot; activating sensor components to monitor the collaborative control process of the tipper operation to obtain real-time tipper information; and dynamically adjusting the collaborative task planning based on the real-time tipper information according to a predetermined collaborative strategy. This application solves the technical problem in existing technologies where the lack of a fine-grained collaborative planning mechanism for tipper operations leads to serial operation of multiple robots, rigid scheduling, and low task efficiency during tipper operations. It achieves the technical effect of intelligent task allocation and dynamic collaborative control of robots, improving the efficiency of tipper operations.

Description

Technical Field

[0001] This application relates to the field of robot control technology, specifically to a collaborative control method and system for a tipper robot used on a train operation line. Background Technology

[0002] In railway transportation systems, tipplers are key equipment for unloading bulk materials and are widely used on train operation lines. With the development of automation technology, robotic systems, such as uncoupling robots, coupling robots, and coupler robots, have been gradually introduced into tippler operations to replace manual labor in completing heavy and dangerous tasks such as coupler assembly and disassembly.

[0003] However, existing control methods for tipper robots generally suffer from problems such as high structural rigidity, weak coordination capabilities, and inflexible response. Most systems employ preset static operation procedures, lacking state awareness of the tipping operation, leading to unreasonable robot task allocation and an inability to flexibly adjust operation plans based on the actual state of the work object (such as train formation and coupler type). Furthermore, the level of coordination between robots is low, with most operations being executed sequentially, resulting in time redundancy and overall low efficiency. Once deviations or anomalies occur during operation, the system struggles to adjust task planning in a timely manner, relying on manual intervention, which affects the continuity and intelligence level of the operation. Summary of the Invention

[0004] This application provides a collaborative control method and system for tipper robots on train operation lines. It solves the technical problems of existing technologies, which lack a fine-grained collaborative planning mechanism for tipper operations, resulting in multiple robots operating sequentially, rigid scheduling, and low task efficiency during tipper operations. It achieves the technical effect of realizing intelligent allocation and dynamic collaborative control of robot tasks, thereby improving the efficiency of tipper operations.

[0005] In view of the above problems, on the one hand, this application provides a collaborative control method for a tipper robot on a train operation line. The method includes: collecting multi-dimensional features of the task to be tipped to obtain a task feature set; performing task allocation analysis of the tipper robot with the task feature set as constraints to obtain a task collaborative plan, wherein the task collaborative plan includes the unhooking plan of the unhooking robot, the forward hooking plan of the forward hooking robot, and the re-hooking plan of the re-hooking robot; activating sensor components to monitor the collaborative control process of the task to be tipped based on the task collaborative plan to obtain real-time tipping information; reading a predetermined collaborative strategy, and dynamically adjusting the task collaborative plan based on the real-time tipping information according to the predetermined collaborative strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the re-hooking plan.

[0006] Preferably, the task allocation analysis of the tipping robot is performed using the task feature set as constraints to obtain task collaborative planning, including: extracting a list of trains in the task to be tipped, the train list including multiple trains with model and location identifiers; performing cluster analysis on multiple trains using the model and location as clustering constraints to obtain clustering results, the clustering results including a first cluster; analyzing the first information of the first train in the first cluster to obtain the first priority of the first cluster; generating a collaborative operation order for the task to be tipped based on the first priority; analyzing the collaborative operation order to obtain the collaborative operation path of the tipping robot, and combining the collaborative operation order with the collaborative operation order to form the task collaborative planning.

[0007] Preferably, the first information includes the first model, first location, and first rollover limit of the first train. Analyzing the first information of the first train in the first cluster to obtain the first priority of the first cluster includes: matching the first unhooking time of the first model; combining the real-time location of the unhooking robot with the first location to obtain the second unhooking time; summing the first unhooking time and the second unhooking time to obtain the predicted unhooking time, and combining it with the first rollover limit to obtain the first priority index; taking the maximum value of the first priority index as the first priority of the first cluster.

[0008] Preferably, the collaborative operation path of the tipper robot is obtained by analyzing the collaborative operation sequence, and the collaborative operation sequence is combined with the collaborative operation sequence to form the task collaborative planning, including: generating the hook removal collaborative operation sequence of the unhooking robot in the tipper operation according to the first priority; analyzing the hook removal collaborative operation sequence to obtain the hook removal collaborative operation path of the unhooking robot, and combining the hook removal collaborative operation sequence to form the hook removal plan; generating the forward hook plan of the forward hook robot with the predicted hook removal time as a constraint; obtaining the predicted forward hook time of the forward hook robot, and combining it with the forward hook plan to obtain the re-hook plan of the re-hook robot; the hook removal plan, the forward hook plan and the re-hook plan constitute the task collaborative planning.

[0009] Preferably, the predetermined coordination strategy is read, and the task coordination plan is dynamically adjusted based on the real-time rollover information according to the predetermined coordination strategy, including: dynamically monitoring the unhooking robot in the task coordination plan through the sensor components to obtain real-time unhooking information; analyzing the real-time unhooking position and real-time unhooking status in the real-time unhooking information to obtain real-time unhooking progress; comparing the real-time unhooking progress with the real-time predetermined unhooking progress in the task coordination plan to obtain real-time unhooking deviation; if the real-time unhooking deviation reaches the predetermined deviation constraint, the positive hook plan and the negative hook plan are dynamically adjusted in sequence.

[0010] Preferably, if the real-time unhooking deviation reaches a predetermined deviation constraint, the positive hook planning and the rehook planning are dynamically adjusted sequentially, further comprising: acquiring real-time operating status parameters of the unhooking robot through the sensor components; acquiring predetermined operating status parameters of the unhooking robot and comparing them with the real-time operating status parameters to obtain a real-time unhooking operation health level; obtaining a real-time unhooking operation anomaly level based on the real-time unhooking operation health level, and using the real-time unhooking operation anomaly level as a weight to dynamically adjust the unhooking planning, the positive hook planning, and the rehook planning sequentially.

[0011] Preferably, after comparing the real-time unhooking progress with the real-time predetermined unhooking progress in the task collaborative planning to obtain the real-time unhooking deviation, the method further includes: if the real-time unhooking deviation does not reach the predetermined deviation constraint, then obtaining the real-time positive hook deviation of the positive hook robot; if the real-time positive hook deviation reaches the predetermined deviation constraint, then dynamically adjusting the re-hooking plan.

[0012] Preferably, after obtaining the real-time hook deviation of the hook robot if the real-time hooking deviation does not reach the predetermined deviation constraint, the method further includes: if the real-time hooking deviation does not reach the predetermined deviation constraint, obtaining the real-time hooking operation anomaly degree of the hook robot; and using the real-time hooking operation anomaly degree as a weight, dynamically adjusting the hooking plan and the re-hooking plan in sequence.

[0013] Preferably, the method further includes: acquiring real-time re-hook deviation, and dynamically adjusting the re-hook plan when the real-time re-hook deviation reaches the predetermined deviation constraint.

[0014] On the other hand, this application also provides a collaborative control system for a tipper robot on a train operation line. The system includes: a task feature collection module for collecting multi-dimensional features of the tipper operation to obtain a task feature set; a task collaborative planning module for performing task allocation analysis of the tipper robot based on the task feature set to obtain a task collaborative plan, wherein the task collaborative plan includes the unhooking plan of the unhooking robot, the forward hooking plan of the forward hooking robot, and the re-hooking plan of the re-hooking robot; a collaborative control monitoring module for activating sensor components to monitor the collaborative control process of the tipper operation based on the task collaborative plan to obtain real-time tipper information; and a collaborative strategy adjustment module for reading a predetermined collaborative strategy and dynamically adjusting the task collaborative plan based on the real-time tipper information according to the predetermined collaborative strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the re-hooking plan.

[0015] One or more technical solutions provided in this application have at least the following beneficial effects:

[0016] By collecting multi-dimensional features of the rollover operation, a comprehensive task feature set is formed, providing a basis for subsequent task planning and ensuring that task allocation is targeted and adaptable. Through task allocation analysis of the rollover robots constrained by the task feature set, corresponding operation plans are formulated for the unhooking robot, the forward hooking robot, and the re-hooking robot, resulting in task coordination planning. This ensures reasonable division of labor and cooperation among the robots in space and time, achieving parallel and efficient collaboration. By activating sensor components to monitor the coordinated control process, real-time rollover information is obtained, constructing a perception feedback mechanism to provide data support for dynamic adjustments and enhance operational controllability. By reading the predetermined coordination strategy and dynamically adjusting the task coordination plan based on real-time rollover information, the task allocation and operation process among robots are optimized to adapt to actual changes.

[0017] In summary, this application establishes a multi-robot collaborative planning mechanism based on task characteristics and combines real-time perception and dynamic feedback control to achieve intelligent division of labor, efficient collaboration, and flexible adjustment of robot roles in rollover operations, significantly improving the automation level, execution efficiency, and operational stability of the rollover operation system.

[0018] 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

[0019] Figure 1 This is a flowchart illustrating the collaborative control method for a tipper robot on a train operation line, as provided in an embodiment of this application.

[0020] Figure 2 This is a flowchart illustrating the task collaborative planning process in the collaborative control method for a tipper robot on a train operation line provided in an embodiment of this application.

[0021] Figure 3 This is a schematic diagram of the structure of the collaborative control system for a tipper robot on a train operation line provided in an embodiment of this application.

[0022] Figure labeling: Task feature collection module 10, Task collaborative planning module 20, Collaborative control and monitoring module 30, Collaborative strategy adjustment module 40. Detailed Implementation

[0023] This application provides a collaborative control method and system for tipper robots on train operation lines, which solves the technical problem in the prior art that the lack of a fine-grained collaborative planning mechanism for tipper operations leads to serial operation of multiple robots, rigid scheduling, and low task efficiency in tipper operations. It achieves the technical effect of realizing intelligent allocation and dynamic collaborative control of robot tasks and improving the efficiency of tipper operations.

[0024] Example 1, as Figure 1 As shown in the embodiment of this application, a collaborative control method for a tipper robot on a train operation line is provided, the method comprising:

[0025] Step S100: Collect multi-dimensional features of the rollover operation to obtain the task feature set.

[0026] Specifically, a derailment operation refers to a train operation that requires unloading, typically involving multiple train units. Multidimensional features refer to various attribute dimensions used to describe the derailment operation's status, encompassing vehicle model, coupler type, location, derailment deadline (time limit), historical operation data, weather and track conditions, etc. The task feature set is a structured data set compiled from these multidimensional features, used to guide subsequent robot task planning.

[0027] By deploying sensing devices (such as high-definition cameras, LiDAR, RFID readers, and geomagnetic sensors) around the train operation line, comprehensive data is collected on the trains to be derailed. This data identifies the location of each train, extracts the vehicle model, and determines the type of coupler used (rotary coupler, semi-automatic coupler, etc.). Furthermore, the derailment deadline (i.e., the derailment limit) for each train in the scheduling plan is obtained. Simultaneously, the current position and operational capabilities of the uncoupling, coupling, and recoupling robots can be read. All this information is integrated into a set of structured data, forming a task feature set, providing input for task planning in subsequent steps.

[0028] Step S200: Perform task allocation analysis for the tipper robot using the task feature set as constraints to obtain task collaborative planning, wherein the task collaborative planning includes hook removal planning for the hook removal robot, hook forward planning for the hook forward robot, and hook re-hook planning for the hook re-hook robot.

[0029] Specifically, task allocation analysis refers to the process of dividing different types of robots (unhooking, forward hooking, and re-hooking) into specific tasks based on task feature sets. Task coordination planning refers to the overall scheduling plan that includes the collaborative operation paths and operation sequences of multiple robots. Unhooking planning, forward hooking planning, and re-hooking planning correspond to the operation sequence and collaborative operation path design of the unhooking robot, forward hooking robot, and re-hooking robot in the current operation task, respectively.

[0030] Using the task feature set obtained in step S100 as constraints, intelligent allocation analysis is performed. First, all coupler operation units that need to be processed are identified. Then, the trains are clustered or ranked based on features such as train type, location, coupler status, and derailment priority. For example, trains with earlier derailment deadlines and shorter estimated operation times are prioritized. Subsequently, combining the current position of the uncoupling robot and historical operation efficiency, the execution sequence and path planning of the uncoupling task are generated, and the operation plans of the main coupling and recoupling robots are derived in sequence, so that a complete closed loop of time coordination and continuous operation is formed among the three robots.

[0031] By using task feature sets as constraints for multi-robot task collaborative planning, intelligent division of labor and operation coordination among robots are realized, solving the efficiency bottleneck caused by traditional serial operation and improving the overall collaborative intelligence level of robots and the throughput capacity of the operating system.

[0032] Step S300: Activate the sensor components to monitor the collaborative control process of the rollover operation based on the task collaborative planning, and obtain real-time rollover information.

[0033] Specifically, sensor components refer to hardware units used to monitor the robot's status and the conditions of the work site, including position sensors, vision recognition modules, motor status monitors, etc. The collaborative control process refers to the operational flow in which multiple robots work synchronously according to a preset task plan. Real-time rollover information refers to the operational status feedback dynamically collected by sensor components, including current position, work progress, execution deviation, fault alarms, etc.

[0034] While performing collaborative task planning, the system activates various sensor components deployed on the robot itself and at the work site to achieve real-time perception of the operational status. The interactive robot control system acquires information such as the position coordinates, operation progress, and execution status of each robot during operation. Simultaneously, it collects visual or data feedback from the on-site environment (such as changes in the coupler status and the availability of the work line) to determine the accuracy, timeliness, and abnormal states of the collaborative operation. This information is integrated into real-time rollover information for subsequent dynamic adjustments. For example, during the execution of the unhooking robot, if its motor torque sensor detects abnormal hook resistance and the vision module identifies that the hook has not completely disengaged, the control system generates a "unhooking failure" alarm and records its current position and abnormal parameters as real-time rollover information.

[0035] Step S400: Read the predetermined collaboration strategy, and dynamically adjust the task collaboration plan based on the real-time rollover information according to the predetermined collaboration strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the re-hooking plan.

[0036] Specifically, the pre-defined collaborative strategy refers to the rules and methods set in advance for the collaborative work of the tipping robot, including task adjustment strategies, priority allocation strategies, and exception handling strategies under different conditions. These strategies guide adjustments to the task collaboration plan when real-time tipping information changes. Dynamic adjustment refers to online modifications made to the original task collaboration plan during operation based on real-time feedback information. Dynamic adjustments to the hook removal plan, hook forwarding plan, and hook re-hooking plan refer to the real-time optimization and updates of the robot's operation path, work sequence, or participating roles in these three stages.

[0037] Once real-time rollover information is fed back to the control system, the pre-configured collaborative strategy is immediately invoked, comparing the actual operational status with the original plan. For example, if there is a delay at a certain task point for the unhooking robot, the impact of the delay on the forward and re-hooking operations will be analyzed, and the operation sequence and timing of subsequent robots will be adjusted accordingly, or even the task will be handed over to a backup robot. Dynamic adjustment is applicable not only to schedule deviations but also to situations such as abnormal robot health or failed actions. The adjustment process specifically includes recalculating and planning parameters such as the unhooking sequence and time interval in the unhooking plan, the forward hooking force and positional accuracy in the forward hooking plan, and the re-hooking timing in the re-hooking plan. The adjusted task collaborative plan will be reissued to each robot, guiding them to continue executing the task according to the new plan, ensuring that the entire rollover operation can adapt to changes in the actual situation and successfully achieve the predetermined goal.

[0038] Further, such as Figure 2 As shown, step S200 includes:

[0039] Step S210: Extract the list of trains to be overturned, the list of trains including multiple trains with model and location identifiers.

[0040] Step S220: Perform cluster analysis on multiple trains using the model and location as clustering constraints to obtain clustering results, which include a first cluster.

[0041] Step S230: Analyze the first information of the first train in the first cluster to obtain the first priority of the first cluster.

[0042] Step S240: Generate the collaborative operation sequence of the vehicle overturning operation according to the first priority.

[0043] Step S250: Analyze the collaborative operation sequence to obtain the collaborative operation path of the tipper robot, and combine it with the collaborative operation sequence to form the task collaborative planning.

[0044] Furthermore, the first information includes the first model, first location, and first rollover limit of the first train.

[0045] Specifically, the train list refers to the collection of information on all trains involved in the current derailment operation. The model identifier indicates the train's structure and performance category, reflecting the complexity of the derailment handling. The location identifier represents the actual coordinates of the train on the work line, such as track station location and train number. Based on the multidimensional feature data obtained in step S100, the structural information of the current operation object is extracted to generate a train list containing all trains. This list not only records the model of each train but also includes their physical locations on the track. This information will subsequently be used as key features for clustering and scheduling analysis.

[0046] Clustering constraints are the grouping criteria used in cluster analysis; in this case, they are the train model and location. A first cluster refers to any one of several subsets derived from the clustering results, containing a group of adjacent trains of similar type. Using clustering algorithms (such as K-Means, DBSCAN, or density-based hierarchical clustering), with train model and location as input features, cluster analysis is performed to group trains with adjacent locations and the same model into the same group, resulting in a clustering result containing multiple first clusters. This clustering result helps integrate similar work objects into a unified scheduling unit, improving collaborative work efficiency.

[0047] The "first train" refers to any train in the first cluster. "First information" refers to the key characteristics of the corresponding first train, including its model, location, and overturning deadline (the final time limit for completing the task). "First priority" assigns an execution priority level to the first cluster, typically ranking higher for more urgent tasks. Selecting any train in the first cluster, its model (first model), location information (first location), and overturning deadline (first overturning deadline) are extracted. Combined with the current state of the unhooking robot (e.g., location, execution efficiency), the time cost required for the task is estimated. This estimate is then compared to the overturning deadline to generate a priority index representing the urgency of the task. A higher priority index indicates that the task should be processed with greater priority.

[0048] Collaborative task order refers to the robot's task sequence plan generated based on the priority of each cluster, i.e., which tasks are performed first and which are performed later. The first priority of all clusters in the first cluster is summarized and used as the sorting basis to generate the collaborative task order. This order will serve as the basis for subsequent robot path planning, enabling the robot to complete the tasks of multiple clusters sequentially according to their urgency.

[0049] A collaborative work path refers to the movement route and operation sequence of a robot on a work line according to the work order. Based on the collaborative work order, and combined with each robot's current position, operational capabilities, and on-site working conditions, path planning algorithms (such as A-STAR and Dijkstra's algorithms) are used to calculate the optimal path for the robot from its current position to the target work position. Simultaneously, the collaborative requirements between robots are considered, such as avoiding path intersections and collisions, and ensuring the continuity of the work. Combining the planned collaborative work paths with the collaborative work order forms a work blueprint containing multiple robot orders and path scheduling information, i.e., task collaborative planning. This blueprint clarifies the specific operation steps, execution time, and movement path of each robot in each task, providing comprehensive guidance for the actual control and execution of the robots.

[0050] Furthermore, step S230 includes:

[0051] Step S231: Match the first unhooking time of the first model.

[0052] Step S232: Combine the real-time position of the unhooking robot with the first position to obtain the second unhooking time.

[0053] Step S233: Sum the first unhooking time and the second unhooking time to obtain the predicted unhooking time, and combine it with the first overturning limit to obtain the first priority index.

[0054] Step S234: Take the maximum value among the first priority indices as the first priority of the first cluster.

[0055] Specifically, the first uncoupling time refers to the time required for a typical uncoupling operation, standardized and matched according to the train model. The train model determines the coupler structure, fastening method, etc., and is a key parameter affecting the complexity of the uncoupling operation. First, the model of the first train in the first cluster (i.e., the first model) is identified. Then, the standard uncoupling time corresponding to this model is searched from the preset model-operation time database and recorded as the first uncoupling time. This first uncoupling time is used to subsequently merge with the dynamic path time to form a more comprehensive task duration estimate.

[0056] The second unhooking time refers to the travel time required for the unhooking robot to move from its current position to the first train. The real-time position of the unhooking robot is obtained through sensor components or a positioning module, and then the time required to reach the location of the first train is calculated. This time depends on factors such as the robot's current coordinates, operating speed, and path accessibility. Combining the train's position coordinates, the shortest path algorithm (such as A-STAR or Dijkstra's algorithm) can be used to calculate the travel distance, and the travel time can be estimated by combining the current speed to obtain the second unhooking time.

[0057] The predicted unhooking time is an estimate of the total time required for the unhooking robot to complete the target unhooking task. The predicted unhooking time is obtained by adding the first and second unhooking times. This predicted time is then compared to the first rollover limit to calculate the first priority index. This first priority index is a quantitative indicator based on the urgency of the task; a larger value indicates a higher priority. An example of calculating the first priority index is as follows: Subtract the predicted unhooking time from the first rollover limit to obtain the remaining buffer time. The smaller this time, the more urgent the task. The remaining buffer time is normalized and then inverted to obtain a positive priority index. For example: S = 1 / (1 + ΔT, ΔT = T1 - T2). Where S is the first priority index, ΔT is the remaining buffer time, T1 is the first rollover limit, and T2 is the predicted unhooking time.

[0058] After calculating the first priority index of each train in the first cluster, these first priority indices are compared, and the maximum value is found. This maximum value is the first priority of that cluster. In this way, in subsequent overall task planning, the tasks of different clusters can be sorted and arranged according to the first priority of each first cluster, prioritizing the first cluster with higher priority, and ensuring the risk sensitivity of task scheduling.

[0059] Furthermore, step S250 includes:

[0060] Step S251: Generate the unhooking cooperative operation sequence of the unhooking robot in the operation to be overturned according to the first priority.

[0061] Step S252: Analyze the unhooking collaborative operation sequence to obtain the unhooking collaborative operation path of the unhooking robot, and combine it with the unhooking collaborative operation sequence to form the unhooking plan.

[0062] Step S253: Using the predicted unhooking time as a constraint, generate the positive hook plan for the positive hook robot.

[0063] Step S254: Obtain the predicted hooking time of the hooking robot and combine it with the hooking plan to obtain the hooking plan of the double hooking robot.

[0064] Step S255: The unhooking plan, the forward hook plan, and the rehook plan constitute the task collaborative plan.

[0065] Specifically, the unhooking collaborative operation sequence refers to the order of operations formed by sorting the tasks to be performed on the overturned vehicle according to the priority of different clusters. After evaluating the priority of multiple first clusters, these first clusters are sorted from high to low priority to form the order in which the unhooking robot needs to execute the tasks. The sorting result is the collaborative operation sequence, which serves as the main framework for the task scheduling of the unhooking robot.

[0066] Based on the established order of the unhooking and decoupling collaborative operations, the system utilizes track topology data, train position coordinates, traffic rules, and the current position of the unhooking robot. It then employs path planning algorithms (such as A-STAR and Dijkstra's algorithms) to generate the shortest or optimal route for the robot. This path, combined with the task sequence, forms an executable unhooking plan, which is then sent to the robot control system.

[0067] Since the hooking operation depends on the unhooking process, the predicted unhooking time is used as a time constraint to set the earliest possible start time for the hooking robot. Based on this time window and the track resource status, combined with the list of trains to be hooked, a similar priority evaluation and path planning algorithm is used to determine the hooking task order and path for the hooking robot, thus obtaining the hooking plan.

[0068] The predicted hooking time is the estimated time required for the hooking robot to complete the hooking task. After hooking planning is completed, the predicted hooking time for each task is estimated based on the number of tasks, task distribution, and historical data of the hooking robot. Subsequently, these predicted hooking times are used as time constraints, combined with train re-coupling requirements and path status, to arrange a specific work sequence and travel path for the re-coupling robot. This path will avoid the working area of ​​the hooking robot. This process ensures that the re-coupling task can start in an orderly manner immediately after the completion of the hooking task, achieving seamless connection of processes. Finally, the generated uncoupling plan, hooking plan, and re-coupling plan are integrated into a unified task coordination plan to build a unified task control framework. This task coordination plan not only includes task sequence, path arrangement, and time synchronization information, but can also embed dynamic adjustment interfaces for real-time monitoring and feedback for updates. This task coordination plan is distributed to each robot via network or control bus to achieve collaborative operation.

[0069] Furthermore, step S400 includes:

[0070] Step S410: The sensor components are used to dynamically monitor the unhooking robot in the task collaborative planning to obtain real-time unhooking information.

[0071] Step S420: Analyze the real-time unhooking position and real-time unhooking status in the real-time unhooking information to obtain the real-time unhooking progress.

[0072] Step S430: Compare the real-time unhooking progress with the real-time predetermined unhooking progress in the task collaborative planning to obtain the real-time unhooking deviation.

[0073] Step S440: If the real-time unhooking deviation reaches the predetermined deviation constraint, the positive hook planning and the re-hooking planning are dynamically adjusted in sequence.

[0074] Specifically, during the rollover operation, sensor components are continuously activated to collect data on the movement trajectory and actions of the unhooking robot. The raw data generated by the sensors is processed by edge computing devices or the main control server, denoised, and key features are extracted to form a standardized real-time unhooking information package, including position coordinates, current operation stage, execution speed, and load status, which serves as the input for subsequent task adjustments. The information acquisition cycle is typically in the millisecond range, ensuring continuous monitoring and timely response.

[0075] Real-time unhooking progress is a quantitative representation of the current execution status of the unhooking robot's task, described as a percentage or stage indicator. Upon receiving real-time unhooking information, the robot's current physical position and operational stage (e.g., approach, alignment, execution, withdrawal) are extracted and matched with the standard stages in the task planning to calculate the current progress. For example, if the path from the task start point to the end point is divided into several sub-stages, the current stage as a percentage of the total task is the real-time progress. Furthermore, progress analysis can incorporate action timing information to enhance the accuracy of judging the execution rhythm. For instance, if the unhooking task consists of 5 sub-stages, and the current execution is at stage 3 with the expected execution time, the output real-time unhooking progress is 60%.

[0076] Based on the schedule set in the task collaboration planning, the expected unhooking progress at the current moment is retrieved and compared with the actual unhooking progress collected in real time to calculate the difference. This difference is the real-time unhooking deviation, which can be positive (lagging), negative (ahead of schedule), or zero (synchronized). A preset deviation constraint threshold, i.e., a predetermined deviation constraint (e.g., ±10%), is set to trigger anomaly detection and strategy adjustment. When the real-time unhooking deviation of the unhooking task is detected to exceed the predetermined deviation constraint, the dynamic adjustment mechanism is immediately triggered. First, the current plan of the positive hook robot is delayed or replanned to adapt to the new time window, and the path is re-optimized to avoid resource conflicts or waiting for idle time. Then, the re-hooking plan is adjusted to ensure that the overall process is not amplified by local anomalies. At the same time, if the unhooking is severely delayed, a backup robot can be scheduled or an emergency strategy (such as skipping some tasks, parallel execution, etc.) can be activated to ensure the smoothness of the entire tipping operation process.

[0077] Furthermore, step S440 also includes:

[0078] Step S441: Obtain the real-time operating status parameters of the unhooking robot through the sensor component.

[0079] Step S442: Obtain the predetermined operating status parameters of the unhooking robot and compare them with the real-time operating status parameters to obtain the real-time unhooking operation health status.

[0080] Step S443: Based on the real-time unhooking operation health status, obtain the real-time unhooking operation anomaly status, and use the real-time unhooking operation anomaly status as a weight to dynamically adjust the unhooking plan, the positive hook plan, and the re-hooking plan in sequence.

[0081] Specifically, real-time operational status parameters refer to key operational indicators of the unhooking robot during its current operation, such as motor current, joint temperature, and vibration intensity. Sensor components here mainly include vibration sensors, current sensors, and temperature sensors, installed on critical parts of the robot. While the unhooking robot performs its unhooking task, its operational status parameters are periodically collected. The collection frequency can be set to several times per second to ensure the real-time nature and completeness of the status data. This data is then filtered and normalized by edge computing devices to form real-time operational status parameters for health diagnosis.

[0082] The predetermined operating status parameters refer to the reference values ​​of the unhooking robot's operating indicators under normal working conditions, which can be obtained through offline modeling, experimental sampling, or historical data statistics. Operating health is an evaluation index of the similarity between the current operating state and the predetermined state, represented by a normalized value (e.g., 0 to 1), with higher values ​​indicating more normal operation. The operating baseline parameters of this model of unhooking robot under similar tasks and environments are retrieved, including joint temperature, motor current, and vibration intensity. Then, the real-time operating parameters are compared one by one with the corresponding predetermined operating status parameters, the deviation of each parameter is calculated, and the current real-time unhooking operating health is output according to a preset health evaluation function (e.g., weighted average, maximum deviation, etc.) to help determine whether a malfunction or efficiency reduction may occur in the current operation.

[0083] The real-time unhooking operation anomaly score is the inverse indicator of the real-time unhooking operation health score, and can be defined as: Anomaly Score = 1 - Health Score. This score is used to assess whether the robot is in an abnormal trend and its severity. The real-time unhooking operation health score is converted into the real-time unhooking operation anomaly score, and then, using the real-time unhooking operation anomaly score as a weight, the unhooking planning, forward hook planning, and re-hooking planning are dynamically adjusted. Specific adjustment methods include: increasing the task interval time of the unhooking robot based on the real-time unhooking operation anomaly score, optimizing task paths to avoid potential failure points, and reserving more buffer time for the forward hooking and re-hooking robots. For example, if the real-time unhooking operation anomaly score is high, the task time of the unhooking robot at each work point can be extended, while adjusting the forward hooking and re-hooking robot plans to correspondingly extend their work times and optimize their paths to reduce waiting time or avoid conflicts. Task replanning is automatically generated based on the current scheduling graph and path planning algorithm, enhancing adaptability and fault tolerance to equipment failures, reducing cascading problems caused by single-point anomalies, and ensuring the continuous and stable operation of multi-robot systems under complex working conditions.

[0084] Furthermore, step S430 includes the following:

[0085] If the real-time unhooking deviation does not reach the predetermined deviation constraint, the real-time hooking deviation of the hooking robot is obtained; if the real-time hooking deviation reaches the predetermined deviation constraint, the re-hooking plan is dynamically adjusted.

[0086] Specifically, when the real-time unhooking deviation does not reach the predetermined deviation constraint, meaning the actual operation progress of the unhooking robot is basically consistent with expectations, the execution status of the hooking robot continues to be monitored. Information such as the real-time position, operating status parameters, and task progress of the hooking robot is obtained through sensor components. The real-time hooking task progress is compared with the predetermined hooking progress in the task coordination plan to calculate the real-time hooking deviation. The calculation process is similar to that for real-time unhooking deviation and will not be elaborated here. If the real-time hooking deviation exceeds the predetermined deviation constraint, the adjustment process of the re-hooking plan is immediately activated to dynamically adjust the re-hooking plan, including re-evaluating the re-hooking start time, obstacle avoidance path, and grasping parameters, to ensure that the re-hooking robot will not misoperate or cause workstation congestion before the hooking is fully ready. The adjustment results are updated to the robot control system. For example, the hooking robot B is scheduled to complete the hooking of the 5th carriage at 10:05, but the sensor feedback indicates that its operation progress is delayed by 1 minute, with a spatial positioning error of 12cm, exceeding the predetermined deviation tolerance. After detecting the deviation, the start time of the re-hooking robot C was delayed from 10:06 to 10:08, and the route was replanned to avoid areas where the hooking operation had not yet been completed, in order to prevent misoperation.

[0087] Furthermore, if the real-time unhooking deviation does not reach the predetermined deviation constraint, the real-time hook deviation of the hooking robot is obtained, and then the process further includes:

[0088] If the real-time hook deviation does not reach the predetermined deviation constraint, the real-time hook operation anomaly degree of the hook robot is obtained; the hook planning and the double hook planning are dynamically adjusted in sequence with the real-time hook operation anomaly degree as the weight.

[0089] Specifically, the real-time hook operation anomaly rate is an indicator that measures the deviation of the hook robot's current operating state from the normal value, reflecting the degree of abnormality in the hook robot's abnormal working state. When the current operation deviation of the hook robot does not exceed the set tolerance, the real-time operating state parameters of the hook robot are retrieved and compared with the preset operating state parameters during normal operation to calculate the real-time hook operation anomaly rate. The specific calculation process can refer to the real-time hook unhooking operation anomaly rate calculation process. Using the real-time hook operation anomaly rate as a weight, the current hook planning is adjusted in an early warning manner, such as reducing the load, extending the operation time, and optimizing the path to avoid high-risk areas; at the same time, the re-hooking planning is also adjusted so that it can flexibly wait for or adapt to the new hook completion state, avoid relying on risky nodes, and improve the stability of the tipping operation process.

[0090] Furthermore, the method also includes:

[0091] The real-time re-hook deviation is obtained, and the re-hook plan is dynamically adjusted when the real-time re-hook deviation reaches the predetermined deviation constraint.

[0092] Specifically, real-time re-hook deviation refers to the difference between the actual progress or status of the re-hook robot during a tipping operation and the re-hook plan in the task coordination plan. During the tipping operation, the re-hook robot is continuously monitored by sensor components, collecting its real-time location information, task progress, and operating status parameters. The real-time re-hook progress is compared with the predetermined re-hook task progress in the task coordination plan to calculate the real-time re-hook deviation. The calculation process can refer to the real-time unhooking deviation calculation process. If the real-time re-hook deviation exceeds the predetermined deviation constraint, it indicates a risk of serious lag, deviation, or execution error in the re-hook task. At this time, a dynamic adjustment mechanism is triggered to optimize the re-hook plan, including but not limited to: recalculating path obstacle avoidance strategies, reallocating the re-hook start time, adjusting the execution sequence, or, if necessary, arranging for a backup robot to intervene, in order to restore collaborative stability and prevent a decrease in system efficiency.

[0093] In summary, the collaborative control method for a tipper robot on a train operation line provided in this application has the following beneficial effects:

[0094] This application proposes a collaborative control method for tipper robots on train operation lines, systematically modeling and dynamically scheduling the efficiency and stability issues in multi-robot collaborative operations. First, by collecting multi-dimensional features of the tipper operation, a task feature set is established, providing a data foundation for subsequent task planning. Then, in the task allocation analysis, the operation units are divided using train model and location as clustering constraints. A collaborative operation order is constructed through priority calculation, and a task collaborative plan is generated by combining the timing and path requirements of various robots for uncoupling, forward hooking, and recoupling, thereby achieving fine-grained parallel task scheduling. During the execution phase, the real-time status of each robot is monitored through sensor components to dynamically acquire actual operation information. Based on a preset collaborative strategy, the actual progress of uncoupling, forward hooking, and recoupling operations is compared with the predetermined progress to determine in real time whether there are deviations. When the deviation exceeds the tolerance threshold, the relevant sub-plans are adjusted according to the anomaly weight, based on the operating status and health, to achieve intelligent correction of the collaborative path. Furthermore, the recoupling task is also incorporated into independent monitoring and dynamic response logic to ensure stable connection of subsequent operation links. Through the above methods, the embodiments of this application effectively solve the problems of rough task arrangement, sequential robot collaboration, and slow abnormal response in the prior art, and significantly improve the collaborative intelligence level, scheduling flexibility and tipping operation efficiency of the tipping robot system.

[0095] Example 2, as Figure 3 As shown, based on the same inventive concept as in Embodiment 1 above, this application provides a collaborative control system for a tipper robot on a train operation line, the system comprising:

[0096] The task feature collection module 10 is used to collect multi-dimensional features of the rollover operation to obtain the task feature set.

[0097] The task collaborative planning module 20 is used to perform task allocation analysis for the tipper robot with the task feature set as constraints, and obtain task collaborative planning. The task collaborative planning includes the hook removal planning of the hook removal robot, the hook forward planning of the hook forward robot, and the hook re-hook planning of the hook re-hook robot.

[0098] The collaborative control monitoring module 30 is used to activate the sensor components to monitor the collaborative control process of the vehicle to be overturned based on the task collaborative planning, and obtain real-time overturning information.

[0099] The collaborative strategy adjustment module 40 is used to read a predetermined collaborative strategy and, based on the real-time rollover information, dynamically adjust the task collaborative planning according to the predetermined collaborative strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the re-hooking plan.

[0100] Furthermore, in this embodiment of the application, the task collaborative planning module 20 is also used to perform the following steps:

[0101] Extract the train list from the derailment operation, the train list includes multiple trains with model and location identifiers; perform cluster analysis on multiple trains using the model and location as clustering constraints to obtain clustering results, the clustering results include a first cluster; analyze the first information of the first train in the first cluster to obtain the first priority of the first cluster; generate the collaborative operation order of the derailment operation according to the first priority; analyze the collaborative operation order to obtain the collaborative operation path of the derailment robot, and combine it with the collaborative operation order to form the task collaborative planning.

[0102] Furthermore, the first information includes the first model, first location, and first rollover limit of the first train. In this embodiment, the task collaborative planning module 20 is also used to perform the following steps:

[0103] Match the first hook removal time of the first model; combine the real-time position of the hook removal robot with the first position to obtain the second hook removal time; sum the first hook removal time and the second hook removal time to obtain the predicted hook removal time, and combine it with the first overturning limit to obtain the first priority index; take the maximum value of the first priority index as the first priority of the first cluster.

[0104] Furthermore, in this embodiment of the application, the task collaborative planning module 20 is also used to perform the following steps:

[0105] The unhooking collaborative operation sequence of the unhooking robot in the derailment operation is generated according to the first priority; the unhooking collaborative operation sequence is analyzed to obtain the unhooking collaborative operation path of the unhooking robot, and the unhooking collaborative operation sequence is combined with the unhooking collaborative operation sequence to form the unhooking plan; the forward hook plan of the forward hook robot is generated with the predicted unhooking time as a constraint; the predicted forward hooking time of the forward hook robot is obtained, and the reverse hooking plan of the reverse hook robot is obtained in combination with the forward hook plan; the unhooking plan, the forward hook plan and the reverse hooking plan constitute the task collaborative planning.

[0106] Furthermore, the module name in this application embodiment is also used to perform the following steps:

[0107] The sensor components are used to dynamically monitor the unhooking robot in the task collaborative planning to obtain real-time unhooking information; the real-time unhooking position and real-time unhooking status in the real-time unhooking information are analyzed to obtain the real-time unhooking progress; the real-time unhooking progress is compared with the real-time predetermined unhooking progress in the task collaborative planning to obtain the real-time unhooking deviation; if the real-time unhooking deviation reaches the predetermined deviation constraint, the positive hook planning and the negative hook planning are dynamically adjusted in sequence.

[0108] Furthermore, in this embodiment of the application, the collaborative strategy adjustment module 40 is also used to perform the following steps:

[0109] The sensor components are used to obtain the real-time operating status parameters of the unhooking robot; the predetermined operating status parameters of the unhooking robot are obtained and compared with the real-time operating status parameters to obtain the real-time unhooking operation health; the real-time unhooking operation anomaly is obtained based on the real-time unhooking operation health, and the unhooking plan, the positive hook plan, and the rehook plan are dynamically adjusted in sequence with the real-time unhooking operation anomaly as the weight.

[0110] Furthermore, in this embodiment of the application, the collaborative strategy adjustment module 40 is also used to perform the following steps:

[0111] If the real-time unhooking deviation does not reach the predetermined deviation constraint, the real-time hooking deviation of the hooking robot is obtained; if the real-time hooking deviation reaches the predetermined deviation constraint, the re-hooking plan is dynamically adjusted.

[0112] Furthermore, in this embodiment of the application, the collaborative strategy adjustment module 40 is also used to perform the following steps:

[0113] If the real-time hook deviation does not reach the predetermined deviation constraint, the real-time hook operation anomaly degree of the hook robot is obtained; the hook planning and the double hook planning are dynamically adjusted in sequence with the real-time hook operation anomaly degree as the weight.

[0114] Furthermore, in this embodiment of the application, the collaborative strategy adjustment module 40 is also used to perform the following steps:

[0115] The real-time re-hook deviation is obtained, and the re-hook plan is dynamically adjusted when the real-time re-hook deviation reaches the predetermined deviation constraint.

[0116] Through the foregoing detailed description of the collaborative control method for a tipper robot used in a train operation line, those skilled in the art can clearly understand that the collaborative control system for a tipper robot used in a train operation line in this embodiment, as corresponding to the system disclosed in Embodiment 2, has corresponding functional modules and beneficial effects as it corresponds to the method disclosed in Embodiment 1. For relevant details, please refer to the description in the method section.

[0117] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A collaborative control method for a tipper robot used on a train operation line, characterized in that, include: Multi-dimensional feature collection is performed on the rollover operation to obtain the task feature set; Task allocation analysis of the tipper robot is performed using the task feature set as constraints to obtain task collaborative planning, wherein the task collaborative planning includes hook removal planning of the hook removal robot, hook forward planning of the hook forward robot, and hook re-hook planning of the hook re-hook robot. The sensor components are activated to monitor the collaborative control process of the rollover operation based on the task collaborative planning, and real-time rollover information is obtained. Read the predetermined collaboration strategy, and dynamically adjust the task collaboration plan based on the real-time rollover information according to the predetermined collaboration strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the rehooking plan; Task allocation analysis for the tipper robot is performed using the aforementioned task feature set as constraints to obtain task collaborative planning, including: Extract the list of trains to be derailed, the list of trains including multiple trains with model and location identifiers; Cluster analysis of multiple trains is performed using the model and location as clustering constraints to obtain clustering results, which include a first cluster. Analyze the first information of the first train in the first cluster to obtain the first priority of the first cluster; The collaborative operation sequence for the rollover operation is generated based on the first priority; The collaborative operation sequence is analyzed to obtain the collaborative operation path of the tipper robot, and the collaborative operation sequence is combined with the collaborative operation sequence to form the task collaborative planning. The first information includes the first model, first location, and first derailment limit of the first train. Analyzing the first information of the first train in the first cluster, the first priority of the first cluster is obtained, including: Match the first unhooking time of the first model; By combining the real-time position of the unhooking robot with the first position, the second unhooking time is obtained; The predicted unhooking time is obtained by summing the first unhooking time and the second unhooking time, and the first priority index is obtained by combining it with the first rollover limit. Take the maximum value among the first priority indices as the first priority of the first cluster; The collaborative operation sequence is analyzed to obtain the collaborative operation path of the tipper robot, and the collaborative operation sequence is combined with the collaborative operation sequence to form the task collaborative planning, including: The unhooking collaborative operation sequence of the unhooking robot in the derailment operation is generated according to the first priority; The unhooking collaborative operation sequence is analyzed to obtain the unhooking collaborative operation path of the unhooking robot, and the unhooking collaborative operation sequence is combined with the unhooking collaborative operation sequence to form the unhooking plan; The forward hook planning of the forward hook robot is generated using the predicted unhooking time as a constraint. Obtain the predicted hooking time of the hooking robot and combine it with the hooking plan to obtain the hooking plan of the double hooking robot; The unhooking plan, the forward hooking plan, and the rehooking plan constitute the task collaborative planning.

2. The collaborative control method for a tipper robot on a train operation line as described in claim 1, characterized in that, Reading a predetermined collaboration strategy and dynamically adjusting the task collaboration plan based on the real-time failure information according to the predetermined collaboration strategy includes: The sensor components are used to dynamically monitor the unhooking robot in the task collaborative planning to obtain real-time unhooking information. Analyze the real-time unhooking position and real-time unhooking status in the real-time unhooking information to obtain the real-time unhooking progress; The real-time unhooking progress is compared with the real-time predetermined unhooking progress in the task collaborative planning to obtain the real-time unhooking deviation. If the real-time unhooking deviation reaches the predetermined deviation constraint, the positive hook planning and the re-hooking planning will be dynamically adjusted in sequence.

3. The collaborative control method for a tipper robot on a train operation line as described in claim 2, characterized in that, If the real-time unhooking deviation reaches the predetermined deviation constraint, the positive hook planning and the re-hooking planning are dynamically adjusted sequentially, and the process further includes: The sensor components are used to acquire the real-time operating status parameters of the unhooking robot. Obtain the predetermined operating status parameters of the unhooking robot and compare them with the real-time operating status parameters to obtain the real-time unhooking operation health status; Based on the real-time unhooking operation health status, the real-time unhooking operation anomaly status is obtained, and the unhooking plan, the positive hook plan, and the rehooking plan are dynamically adjusted in sequence using the real-time unhooking operation anomaly status as a weight.

4. The collaborative control method for a tipper robot on a train operation line as described in claim 3, characterized in that, After comparing the real-time unhooking progress with the real-time predetermined unhooking progress in the task collaborative planning to obtain the real-time unhooking deviation, the method further includes: If the real-time unhooking deviation does not reach the predetermined deviation constraint, then the real-time hook deviation of the hook robot is obtained; If the real-time positive hook deviation reaches the predetermined deviation constraint, the double hook plan will be dynamically adjusted.

5. The collaborative control method for a tipper robot on a train operation line as described in claim 4, characterized in that, If the real-time unhooking deviation does not reach the predetermined deviation constraint, then after obtaining the real-time hook deviation of the hooking robot, the method further includes: If the real-time hook deviation does not reach the predetermined deviation constraint, the real-time hook operation abnormality of the hook robot is obtained; Using the real-time anomaly degree of the positive hook operation as a weight, the positive hook planning and the double hook planning are dynamically adjusted sequentially.

6. The collaborative control method for a tipper robot on a train operation line as described in claim 5, characterized in that, include: The real-time re-hook deviation is obtained, and the re-hook plan is dynamically adjusted when the real-time re-hook deviation reaches the predetermined deviation constraint.

7. A collaborative control system for a tipper robot used on a train operation line, characterized in that, The system is used to execute the collaborative control method for a tipper robot on a train operation line as described in any one of claims 1-6, including: The task feature collection module is used to collect multi-dimensional features of the rollover operation to obtain the task feature set. The task coordination planning module is used to perform task allocation analysis for the tipper robot with the task feature set as constraints, and obtain task coordination planning. The task coordination planning includes the hook removal planning of the hook removal robot, the hook forward planning of the hook forward robot, and the hook re-hook planning of the hook re-hook robot. The collaborative control monitoring module is used to activate the sensor components to monitor the collaborative control process of the vehicle to be overturned based on the task collaborative planning, and obtain real-time overturning information; The collaborative strategy adjustment module is used to read a predetermined collaborative strategy and, based on the real-time rollover information, dynamically adjust the task collaborative planning according to the predetermined collaborative strategy, including dynamically adjusting the unhooking plan and / or the forward hooking plan and / or the re-hooking plan.

Images