Multi-vehicle adaptive hooking robot working condition self-adaptive adjustment method and system

Abstract

The application discloses a kind of multi-vehicle adaptation hooking robot working condition self-adapting adjustment method and system, and the application relates to the technical field of self-adapting adjustment method, according to the position of the hooking feature and the hooking work area of the multi-vehicle adaptation hooking robot, the hooking behavior event of the multi-vehicle adaptation hooking robot relative to the hooking feature is determined, and the adaptive hooking event of the multi-vehicle adaptation hooking robot is determined according to each hooking behavior event and the movement path of the multi-vehicle adaptation hooking robot.Therefore, based on each working condition type and the corresponding adaptive hooking event, the adaptive hooking behavior of the multi-vehicle adaptation hooking robot is determined;Based on the corresponding hooking work efficiency estimated by each adaptive hooking behavior, according to the hooking work efficiency, the corresponding hooking position and the remaining work content of the multi-vehicle adaptation hooking robot, the remaining hooking work system of the multi-vehicle adaptation hooking robot is determined, and the accuracy of the remaining hooking work system is improved.

Description

Technical Field

[0001] This invention relates to the technical field of adaptive adjustment methods, and more particularly to an adaptive adjustment method and system for multi-vehicle adaptive unhooking robot under working conditions. Background Technology

[0002] With the development of technology, multi-vehicle adaptive unhooking robots are robotic systems that can adapt to various vehicle models and automatically complete unhooking operations. They perform corresponding unhooking operations under specific working conditions. In existing technologies, multi-vehicle adaptive unhooking robots perform unhooking at corresponding unhooking positions and around the unhooking work area without considering the robot's movement path. This affects the accuracy of the robot's adaptive unhooking events and lacks corresponding adaptive control over the unhooking behavior, thus impacting the accuracy of the remaining unhooking work system. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a method and system for adaptive adjustment of working conditions for multi-vehicle adaptive unhooking robots.

[0004] This invention provides a method for adaptive adjustment of the working conditions of a multi-vehicle adaptive unhooking robot, comprising: determining multiple region nodes based on the movement path of the multi-vehicle adaptive unhooking robot; acquiring multiple region working images based on the detection of the multiple region nodes; determining multiple environmental features, dynamic features, and unhooking features based on image recognition of each region working image; determining the unhooking working area based on the multiple environmental features, dynamic features, and unhooking features; and determining the unhooking position of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot. The system identifies hooking behavior events and determines adaptive hooking events for the multi-vehicle adaptive hooking robot based on each hooking behavior event and the robot's movement path. It also determines the working condition type of the multi-vehicle adaptive hooking robot based on each hooking location and multiple environmental features, and determines the adaptive hooking behavior based on each working condition type and the corresponding adaptive hooking event. Finally, it estimates the corresponding hooking efficiency based on each adaptive hooking behavior, and determines the remaining hooking work system for the multi-vehicle adaptive hooking robot based on the hooking efficiency, the corresponding hooking location, and the remaining work content of the multi-vehicle adaptive hooking robot.

[0005] This invention provides a multi-vehicle adaptive uncoupling robot working condition adaptive adjustment system, which is applied to the above-mentioned multi-vehicle adaptive uncoupling robot working condition adaptive adjustment method. The multi-vehicle adaptive uncoupling robot working condition adaptive adjustment system includes:

[0006] The regional working image module is used to determine multiple regional nodes based on the movement path of the multi-vehicle adaptive unhooking robot, and to acquire multiple regional working images based on the detection of multiple regional nodes.

[0007] The multimodal data module is used to determine multiple environmental features, dynamic features, and de-hooking features based on image recognition of working images in various regions, and to determine the de-hooking working area based on these features.

[0008] The adaptive unhooking event module is used to determine the unhooking behavior event of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot, and to determine the adaptive unhooking event of the multi-vehicle adaptive unhooking robot based on each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot.

[0009] The adaptive unhooking behavior module is used to determine the working condition type of the multi-vehicle adaptive unhooking robot based on each unhooking position and multiple environmental features, and to determine the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each working condition type and the corresponding adaptive unhooking event.

[0010] The remaining unhooking system module is used to predict the corresponding unhooking efficiency based on each adaptive unhooking behavior, and to determine the remaining unhooking system of the multi-vehicle adaptive unhooking robot based on the unhooking efficiency, the corresponding unhooking position, and the remaining work content of the multi-vehicle adaptive unhooking robot.

[0011] Compared with the prior art, the beneficial effects of the present invention are:

[0012] In this embodiment of the invention, the method determines the unhooking behavior event of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot. The adaptive unhooking event of the multi-vehicle adaptive unhooking robot is then determined based on each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot. The introduction of the unhooking working area, combined with a holistic consideration of each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot, improves the accuracy of the adaptive unhooking event of the multi-vehicle adaptive unhooking robot.

[0013] Therefore, the working conditions of the multi-vehicle adaptive unhooking robot are determined based on each unhooking location and multiple environmental characteristics. The adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot is determined based on each working condition and the corresponding adaptive unhooking event. The corresponding unhooking efficiency is estimated based on each adaptive unhooking behavior. The remaining unhooking work system of the multi-vehicle adaptive unhooking robot is determined based on the unhooking efficiency, the corresponding unhooking location, and the remaining work content of the multi-vehicle adaptive unhooking robot. This introduces the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot, achieving a holistic consideration of unhooking efficiency, the corresponding unhooking location, and the remaining work content of the multi-vehicle adaptive unhooking robot, thus improving the accuracy of the remaining unhooking work system. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the adaptive adjustment method for the working conditions of a multi-vehicle adaptive unhooking robot in an embodiment of the present invention.

[0015] Figure 2 This is a flowchart illustrating step S11 in the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method in an embodiment of the present invention.

[0016] Figure 3 This is a flowchart illustrating step S12 in the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method in an embodiment of the present invention.

[0017] Figure 4 This is a flowchart illustrating step S13 in the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method in an embodiment of the present invention.

[0018] Figure 5 This is a flowchart illustrating step S14 in the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method in an embodiment of the present invention.

[0019] Figure 6 This is a flowchart illustrating step S15 in the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method in an embodiment of the present invention.

[0020] Figure 7 This is a schematic diagram of the structural composition of the multi-vehicle adaptive unhooking robot working condition adaptive adjustment system in an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0022] Please see Figures 1 to 7 A method for adaptive adjustment of working conditions for multi-vehicle adapted unhooking robots, applied to adaptive adjustment scenarios; the method for adaptive adjustment of working conditions for multi-vehicle adapted unhooking robots includes:

[0023] Step S11: Determine multiple regional nodes based on the movement path of the multi-vehicle adaptive unhooking robot, and collect multiple regional working images based on the detection of multiple regional nodes;

[0024] Step S12: Based on image recognition of working images in each region, determine multiple environmental features, dynamic features, and de-hooking features, and determine the de-hooking working area based on the multiple environmental features, dynamic features, and de-hooking features;

[0025] Step S13: Determine the unhooking behavior event of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot. Determine the adaptive unhooking event of the multi-vehicle adaptive unhooking robot based on each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot.

[0026] Step S14: Determine the working condition type of the multi-vehicle adaptive unhooking robot based on each unhooking position and multiple environmental features, and determine the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each working condition type and the corresponding adaptive unhooking event.

[0027] Step S15: Based on the estimated unhooking efficiency of each adaptive unhooking behavior, determine the remaining unhooking system of the multi-vehicle adaptive unhooking robot according to the unhooking efficiency, the corresponding unhooking position, and the remaining work of the multi-vehicle adaptive unhooking robot.

[0028] refer to Figure 2 In step S11, the specific steps are as follows:

[0029] S111: Collect the current position and target position of the multi-vehicle adaptive unhooking robot, determine the movement path of the multi-vehicle adaptive unhooking robot based on the current position, target position and corresponding movement space, determine multiple unhooking operation areas based on the detection of the movement path of the multi-vehicle adaptive unhooking robot, and mark the area nodes of each unhooking operation area.

[0030] S112: In each regional node, multiple cameras are triggered based on the node position of the regional node. The multiple cameras take a circular picture of the unhooking operation area where the regional node is located and collect multiple sub-region images. Based on the multiple sub-region images, the corresponding unhooking position, and the previous positioning position of the multi-vehicle adaptive unhooking robot, the corresponding regional working image is determined.

[0031] In the embodiments of this application, the robot's precise coordinate position is obtained in real time through the robot's built-in positioning system (such as GPS, LiDAR SLAM, or visual positioning system), while the target position is obtained through a task scheduling system or manual input, and usually includes three-dimensional spatial coordinates and posture information.

[0032] The movement path of the multi-vehicle adaptive unhooking robot is determined based on the current position, the target position, and the corresponding movement space. The movement space refers to the three-dimensional space range in which the robot can operate, including physical limitations, safety boundaries, and operational constraints. The path planning algorithm (such as RRT or dynamic window method) will comprehensively consider these factors and calculate the optimal path from the current position to the target position. The planned movement path is divided into multiple continuous working areas according to certain rules (such as distance intervals, functional areas, or environmental characteristics). Each area represents a relatively independent unhooking operation unit. A unique identifier node is set for each working area, containing information such as area ID, spatial coordinates, and range boundaries.

[0033] Furthermore, when the unhooking robot reaches or approaches a preset area node, the system will automatically trigger multiple cameras deployed around that node. The triggering mechanism is based on a distance threshold. Typically, when the distance between the robot and the node is less than a preset value (e.g., 5 meters), the system will activate the camera network. The camera triggering needs to consider the time synchronization issue to ensure that all cameras start collecting data at the same time to avoid the time difference of image data affecting subsequent processing. At the same time, the triggering mechanism needs to have fault tolerance to prevent false triggering or missed triggering.

[0034] Multiple cameras take circular images of the unhooking operation area where the regional node is located. Circular shooting means that multiple cameras simultaneously capture images of the same operation area from different angles (usually 360 degrees full coverage). The cameras are usually distributed around the operation area at fixed intervals (such as 45 degrees or 60 degrees) to form an observation network without blind spots. The camera layout needs to consider overlapping areas to ensure that the fields of view of adjacent cameras have appropriate overlap, which is convenient for subsequent image stitching. At the same time, the focal length and shooting parameters of the cameras need to be adjusted according to the size of the operation area to ensure that the image resolution meets the recognition requirements.

[0035] Each camera is responsible for capturing images of a sub-region within its field of view; these sub-region images are raw image data containing visual information from different angles of the work area; image acquisition is usually in high-resolution format and includes timestamps and camera position information; image acquisition needs to take into account changes in lighting conditions and requires automatic adjustment of exposure parameters; at the same time, the image data volume is usually large, requiring efficient compression and transmission mechanisms to ensure data real-time performance.

[0036] Based on multiple sub-region images, the corresponding unhooking positions, and the previous positioning positions of the multi-vehicle adaptive unhooking robots, the system determines the corresponding regional working image and stitches these sub-region images into a complete regional working image. The system optimizes the image by combining the unhooking position (the specific position where the operation needs to be performed) and the robot's previous positioning positions (historical trajectory data), enhancing the image quality of key areas. Image stitching requires precise coordinate transformation and color correction to ensure seamless stitching. At the same time, the system enhances key areas in the image based on the unhooking position and the robot's historical positions, improving the accuracy of subsequent feature extraction.

[0037] refer to Figure 3 In step S12, the specific steps are as follows:

[0038] S121: In each region working image, multiple sub-feature images are determined based on the recognition of the region working image. The corresponding image recognition mode is determined based on the position, image contour and image type weight of the multiple sub-feature images. Different types of sub-feature images are recognized using the corresponding image recognition mode.

[0039] S122: Execute the corresponding image recognition mode for multiple sub-feature images and output the corresponding feature combination, which has multiple feature combinations in the regional working image;

[0040] S123: Based on the screening of multiple feature combinations, multiple environmental features, dynamic features and de-hooking features are determined. A preliminary region is constructed based on the multiple environmental features and dynamic features. The de-hooking work area is determined based on the preliminary region and the de-hooking features.

[0041] In the embodiments of this application, multiple sub-feature images are determined based on the recognition of each regional working image. The system first preprocesses the regional working images, including image enhancement, noise removal, and contrast adjustment, to improve image quality. The complete regional working image is then segmented into multiple sub-regions with independent features using image segmentation algorithms (such as watershed segmentation, region growing, and edge detection). Each sub-feature image represents a potential object of interest or region, such as a carriage connector, ground markings, or operating equipment. The system assigns a unique identifier to each sub-feature image and records its basic information, such as its position coordinates and size in the original image. The selection of the image segmentation algorithm is based on image characteristics and processing speed requirements, and multiple algorithms are usually combined to improve segmentation accuracy. The minimum size of the sub-feature images is subject to a threshold limit to avoid over-segmentation and wasted computational resources. At the same time, the segmentation process considers multi-dimensional features such as image texture, color, and edges to ensure the semantic integrity of the segmentation results.

[0042] For each sub-feature image, the system analyzes three key parameters: location information, image contour, and image category weight. Location information includes the absolute coordinates and relative position of the sub-feature image within the working image of the region, used to determine the spatial importance of the feature and its correlation with other features. Image contour is obtained through contour extraction algorithms, capturing geometric features such as the boundary shape, aspect ratio, roundness, and complexity of the sub-feature image. These features reflect the basic shape of the target object. Image category weight is achieved by pre-establishing a feature type weight table, where different types of features have different processing priorities and importance coefficients. For example, the weight of the carriage connector feature is 0.9, while the weight of the ground marking feature is 0.6. Based on these three parameters, the system determines the most suitable recognition mode for the sub-feature image using a decision matrix or machine learning model. Recognition modes include: template matching (suitable for fixed-shape features), deep learning recognition (suitable for complex deformation features), edge feature analysis (suitable for features with obvious contours), and texture analysis (suitable for objects with rich surface features).

[0043] After assigning a defined recognition mode to each sub-feature image, the system executes multiple recognition tasks in parallel. Different recognition modes have different input requirements and output formats: Template matching mode: input sub-feature image and template library, output matching score and best matching position; Deep learning mode: input normalized sub-feature image, output classification probability and feature vector; Edge feature analysis: input edge detection result, output geometric parameters; Texture analysis: input gray-level co-occurrence matrix, output texture feature parameters; The recognition results undergo confidence evaluation, and only recognition results that reach a preset threshold are accepted; The execution of recognition modes adopts multi-threaded or distributed computing to improve processing efficiency; The system has mode switching capability, and when the recognition effect of a certain mode is not good, it can automatically try a backup mode; The recognition results contain rich metadata, such as processing time, confidence, feature parameters, etc., for use in subsequent steps.

[0044] Furthermore, the corresponding image recognition mode is executed on multiple sub-feature images, and the corresponding feature combination is output. The regional working image has multiple feature combinations, thus introducing multiple feature combinations.

[0045] At this point, the system calls the corresponding recognition algorithm for processing based on the recognition pattern of each sub-feature image determined in S121. The recognition process includes feature extraction, pattern matching, classification judgment, and other steps, with different recognition patterns having their own specific processing procedures. The system processes multiple sub-feature images in parallel to improve overall processing efficiency while ensuring reasonable resource allocation among recognition tasks. During the execution of each recognition pattern, the system records key information such as processing time, confidence level, and intermediate results for subsequent quality assessment. Parallel processing adopts a multi-threaded or distributed computing architecture, dynamically allocating computing resources according to the complexity of the recognition pattern. The recognition algorithm will adaptively adjust based on historical data, such as fine-tuning the parameters of the deep learning model and optimizing the threshold of template matching.

[0046] After each sub-feature image is recognized, it outputs one or more feature vectors, which contain the key information of that sub-feature. The system combines related feature vectors to form feature combinations. The criteria for feature combination include spatial proximity, functional relevance, and temporal consistency. Each feature combination contains multiple original feature vectors, their relationships, and combination weights. Redundancy checks are performed on feature combinations to remove duplicate or highly related features, ensuring the simplicity and effectiveness of the combination. Feature combination uses graph theory or clustering algorithms to effectively combine spatially adjacent or functionally related features. The timeliness of features is considered during the combination process, prioritizing the most recently acquired feature data. The system calculates a comprehensive confidence score for each feature combination, which serves as an important basis for subsequent screening.

[0047] A complete regional working image typically contains multiple feature combinations, each representing a specific scene element or functional region. The system maintains the spatial and logical relationships between these feature combinations to form a complete scene understanding. Feature combinations undergo spatial consistency verification to ensure their distribution in the regional working image conforms to the physical laws of the actual scene. The system periodically updates feature combinations to adapt to dynamic scene changes and maintain their timeliness. Feature combination management employs spatial indexing techniques, such as R-trees or quadtrees, to improve the efficiency of spatial queries and relationship analysis. The system establishes a semantic network between feature combinations to represent their functional dependencies and logical relationships. The dynamic update mechanism uses an incremental algorithm, updating only the feature combinations that have changed, reducing computational overhead.

[0048] Therefore, multiple environmental features, dynamic features, and de-hooking features are determined by screening multiple feature combinations. A preliminary region is constructed based on the multiple environmental features and dynamic features. The de-hooking working region is determined based on the preliminary region and the de-hooking features. This approach takes into account the overall consideration of the preliminary region and the de-hooking features, ensuring the accuracy of the de-hooking working region.

[0049] At this point, the system first filters and classifies the multiple feature combinations generated in S122. The filtering is based on preset evaluation criteria, including the confidence level, completeness, and timeliness of the feature combinations. Environmental features refer to relatively stable scene elements, such as fixed equipment, ground markings, and building structures. The system determines environmental features by analyzing static elements and spatial relationships within the feature combinations. Dynamic features refer to changing or moving elements in the scene, such as personnel activity, vehicle movement, and changes in lighting. The system identifies dynamic features by comparing changes in feature combinations across consecutive time frames. Unhooking features refer to specific elements directly related to the unhooking operation, such as the carriage connector and latch. The system identifies hook removal features through shape matching, position analysis, and function recognition, including devices and operating handles. Feature selection employs a multi-level filtering mechanism: first, low-quality features are filtered based on confidence thresholds; then, redundant features are removed based on correlation analysis; and finally, key features are retained based on importance assessment. Environmental feature recognition uses spatial consistency verification to ensure that identified environmental features remain stable across multiple image frames. Dynamic feature recognition uses motion detection algorithms, such as optical flow and background subtraction, to analyze the temporal changes in feature combinations. Hook removal feature recognition utilizes a specialized template library and deep learning models for precise matching of different types of connection devices.

[0050] Feature selection employs a multi-level filtering mechanism. First, low-quality features are filtered based on confidence thresholds. Then, redundant features are removed based on correlation analysis. Finally, key features are retained based on importance assessment. Environmental feature recognition uses spatial consistency verification to ensure that the identified environmental features remain stable across multiple frames of images. Dynamic feature recognition uses motion detection algorithms, such as optical flow and background subtraction, to analyze the temporal changes of feature combinations. Hook / unhook feature recognition uses a specialized template library and deep learning models to accurately match different types of connection devices.

[0051] The system analyzes the spatial relationship between each preliminary area and the unhooking feature to determine the most suitable area for unhooking operations. The determination of the unhooking work area comprehensively considers the location distribution of the unhooking feature, the functional attributes of the preliminary area, and operational safety requirements. The system conducts a risk assessment on candidate unhooking work areas, considering factors such as obstacle distribution, personnel activity, and equipment accessibility. The final determined unhooking work area includes precise spatial boundaries, operation entry points, safe evacuation paths, and other key information. The determination of the unhooking work area adopts a multi-objective optimization algorithm to balance operational efficiency, safety, and accessibility.

[0052] refer to Figure 4 In step S13, the specific steps are as follows:

[0053] S131: Collect the de-hooking working area, determine multiple de-hooking working nodes based on the identification of the de-hooking working area, and determine the location event based on the location of the multiple de-hooking working nodes and the location of the de-hooking features;

[0054] S132: Determine action events based on multiple unhooking work nodes and the unhooking work arm of the multi-vehicle-adaptive unhooking robot; determine the unhooking behavior event of the multi-vehicle-adaptive unhooking robot relative to the unhooking feature based on position events, action events, and the working mode of the multi-vehicle-adaptive unhooking robot relative to the unhooking work area.

[0055] S133: Collect the movement path of the multi-vehicle adaptive unhooking robot, determine the activity range of the corresponding unhooking behavior based on the detection of each unhooking behavior event, and determine the adaptive unhooking event of the multi-vehicle adaptive unhooking robot based on the activity range of the unhooking behavior, the movement path of the multi-vehicle adaptive unhooking robot, and the unhooking posture of the multi-vehicle adaptive unhooking robot.

[0056] In the embodiments of this application, the unhooking working area is output by step S123, including the three-dimensional spatial coordinate range, the position of unhooking features (such as connectors, latches, etc.), and environmental structure information (such as obstacles, ground markings, etc.). The system collects images and point cloud data of the unhooking working area in real time through visual sensors (such as cameras, LiDAR), and performs regional modeling by combining historical data. The collected data is converted into a three-dimensional spatial model, and key feature points, boundaries, obstacles, and other information are marked to form a computable spatial structure.

[0057] Within the unhooking work area, the system automatically generates multiple unhooking work nodes based on the spatial layout, distribution of unhooking features, and robot accessibility. These nodes are key spatial locations for the robot to perform operations. Node types include: proximity nodes: the position where the robot approaches the unhooking feature, typically 1-2 meters from the target; alignment nodes: the position where the robot aligns with the unhooking feature, typically 0.5-1 meter from the target; and operation nodes: the position where the robot performs the unhooking operation, typically in direct contact with the target. Spatial sampling, clustering, or feature density-based methods are used to ensure that the node distribution is reasonable and covers all operation positions. Node generation algorithm: Spatial sampling, clustering, or feature density-based methods are used to ensure that the node distribution is reasonable and covers all operation positions.

[0058] Position events are determined based on the locations of multiple unhooking work nodes and the location of the unhooking feature. These position events describe the relative positional relationship between the robot and the unhooking feature and are used to trigger subsequent actions. Examples include: Approach event: the robot reaches the approach node, triggering deceleration and visual alignment; Alignment event: the robot reaches the alignment node, triggering fine-tuning of the outrigger; Operation event: the robot reaches the operation node, triggering the unhooking action. The system calculates the spatial relationship (distance, angle, direction) between each unhooking work node and the unhooking feature, generating the corresponding position event.

[0059] Specifically, the uncoupling work area is a railway freight station. The robot needs to perform uncoupling operations on three different types of carriages (Type A, Type B, and Type C). Uncoupling characteristics: Type A carriage: the connector is located in the middle of the carriage, 1.2 meters high; Type B carriage: the connector is located at the end of the carriage, 1.5 meters high; Type C carriage: the connector is located at the bottom of the carriage, 0.8 meters high. The robot's initial position is at the station coordinates (100, 200), with the arm fully retracted.

[0060] The system collects point cloud and image data of the three carriages using lidar and cameras; generates a three-dimensional region model, and marks the connector positions of carriages A, B, and C and surrounding obstacles (such as signal lights and pillars).

[0061] Determining the uncoupling working nodes: Type A car: Approach node N1: (150, 250), 1.5 meters from the connector; Alignment node N2: (152, 252), 0.8 meters from the connector; Operation node N3: (153, 253), in direct contact with the connector; Type B car: Approach node N4: (200, 300), 1.8 meters from the connector; Alignment node N5: (202, 302), 1.0 meter from the connector; Operation node N6: (203, 303), in direct contact with the connector; Type C car: Approach node N7: (250, 350), 1.2 meters from the connector; Alignment node N8: (252, 352), 0.6 meters from the connector; Operation node N9: (253, 353), in direct contact with the connector.

[0062] Location event generation: Type A Car: Event E1: Robot reaches N1, triggering "Approaching Type A Connector"; Event E2: Robot reaches N2, triggering "Aligning with Type A Connector"; Event E3: Robot reaches N3, triggering "Operating Type A Connector"; Type B Car: Event E4: Robot reaches N4, triggering "Approaching Type B Connector"; Event E5: Robot reaches N5, triggering "Aligning with Type B Connector"; Event E6: Robot reaches N6, triggering "Operating Type B Connector"; Type C Car: Event E7: Robot... When the robot reaches N7, it triggers "Approaching the C-type connector"; Event E8: When the robot reaches N8, it triggers "Aligning with the C-type connector"; Event E9: When the robot reaches N9, it triggers "Operating the C-type connector"; The system generates 9 position events (E1-E9), each event containing: trigger condition (robot reaches the specified node); event type (approach, alignment, operation); associated unhooking feature (A / B / C type connector); These position events will be used as inputs to step S132 to generate motion events and unhooking behavior events.

[0063] Furthermore, action events are determined based on multiple unhooking work nodes and the unhooking work arm of the multi-vehicle-adaptive unhooking robot. Unhooking behavior events of the multi-vehicle-adaptive unhooking robot relative to the unhooking feature are determined based on position events, action events, and the working mode of the multi-vehicle-adaptive unhooking robot relative to the unhooking work area. This approach takes into account the overall consideration of position events, action events, and the working mode of the multi-vehicle-adaptive unhooking robot relative to the unhooking work area, ensuring the accuracy of the unhooking behavior events of the multi-vehicle-adaptive unhooking robot relative to the unhooking feature.

[0064] At this point, multiple unhooking work nodes (such as approach nodes, alignment nodes, and operation nodes) output from step S131 are generated. Each node contains spatial coordinates and associated unhooking features. Based on the robot's arm structure (such as a multi-joint manipulator, linear actuator, etc.) and kinematic model, the system calculates the reachability, attitude, and motion trajectory of the arm at each node. Arm movements include: extension, rotation, clamping, and releasing, and each movement corresponds to specific control parameters (such as angle, speed, and torque).

[0065] For each unhooking operation node, the system generates a corresponding action event, describing the specific action that the outrigger needs to perform at that node. The action event includes: action type (such as "outrigger extension" or "gripper closure"); action parameters (such as extension length and clamping force); triggering conditions (such as "reaching the alignment node"); execution time (such as "action lasts for 2 seconds"); force feedback control is introduced to limit the torque of clamping and releasing actions to avoid damaging the equipment.

[0066] The system determines the uncoupling behavior event based on position events, motion events, and working modes. Position event input: position events generated by S131 (such as "reaching the approach node" or "reaching the operation node"); motion event input: motion events generated in this step (such as "outrigger extension" or "gripper closure"); the system selects the working mode according to the current working conditions (such as vehicle type, environmental conditions, and task priority). For example: Standard mode: suitable for conventional vehicle types, smooth movement, efficiency priority; Fine mode: suitable for precision vehicle types, slow movement, accuracy priority; Emergency mode: suitable for abnormal situations, fast movement, safety priority.

[0067] The system combines position events, motion events, and working modes to generate unhooking behavior events. Unhooking behavior events describe the complete behavioral logic of the robot performing a specific action in a specific position and mode. Behavior events include: behavior type (such as "approach unhooking", "align unhooking", "execute unhooking"); triggering conditions (position event + motion event); working mode (such as "standard mode"); behavior parameters (such as motion speed, torque limit); and subsequent behavior (such as "return to standby point after completion").

[0068] Specifically, the multi-vehicle adaptable unhooking robot needs to perform unhooking operations within the unhooking work areas of three vehicle types (Type A, Type B, and Type C). The unhooking features (connectors) positions and structures differ for each vehicle type, requiring different arm movements. The system adopts two working modes: "standard mode" and "fine mode". Unhooking work node inputs (from S131): Type A vehicle: Approach node N1: (100, 200); Alignment node N2: (101, 201); Operation node N3: (102, 202); Type B vehicle: Approach node N4: (150, 250); Alignment node N5: (151, 251); Operation node N6: (152, 252); Type C vehicle: Approach node N7: (200, 300); Alignment node N8: (201, 301); Operation node N9: (202, 302).

[0069] Action event generation: Type A car: Action event A1: Outrigger extends to 1.2 meters (trigger condition: reaches N1); Action event A2: Outrigger rotates to 30 degrees (trigger condition: reaches N2); Action event A3: Grippers close, torque 10 N·m (trigger condition: reaches N3); Type B car: Action event B1: Outrigger extends to 1.5 meters (trigger condition: reaches N4); Action event B2: Outrigger rotates to 45 degrees (trigger condition: reaches N5); Action event B3: Grippers close, torque 15 N·m (trigger condition: reaches N6); Type C car: Action event C1: Outrigger extends to 1.8 meters (trigger condition: reaches N7); Action event C2: Outrigger rotates to 60 degrees (trigger condition: reaches N8); Action event C3: Grippers close, torque 20 N·m (trigger condition: reaches N9).

[0070] Uncoupling behavior event generation: Type A car (standard mode): Behavior event BE1: Approaching uncoupling (trigger condition: position event E1 + action event A1); Behavior event BE2: Aligning uncoupling (trigger condition: position event E2 + action event A2); Behavior event BE3: Executing uncoupling (trigger condition: position event E3 + action event A3); Type B car (refined mode): Behavior event BE4: Approaching uncoupling (trigger condition: position event E4 + action event B1, action speed reduced by 20%); Behavior event BE5: Aligning uncoupling Hook (Trigger condition: Position event E5 + Action event B2, action speed reduced by 20%); Behavior event BE6: Perform uncoupling (Trigger condition: Position event E6 + Action event B3, torque limit increased by 10%); Type C car (Standard mode): Behavior event BE7: Approach uncoupling (Trigger condition: Position event E7 + Action event C1); Behavior event BE8: Align uncoupling (Trigger condition: Position event E8 + Action event C2); Behavior event BE9: Perform uncoupling (Trigger condition: Position event E9 + Action event C3).

[0071] Output: The system generates 9 unhooking behavior events (BE1-BE9). Each event includes: triggering condition (position event + action event); behavior type (approach, alignment, execution); working mode (standard / fine); behavior parameters (speed, torque, etc.). These behavior events will be used as input to step S133 to determine the adaptive unhooking event.

[0072] Therefore, the movement paths of the multi-vehicle adaptive unhooking robot are collected, and the activity range of the corresponding unhooking behavior is determined based on the detection of each unhooking behavior event. Adaptive unhooking events of the multi-vehicle adaptive unhooking robot are determined based on the activity range of the unhooking behavior, the movement path of the multi-vehicle adaptive unhooking robot, and the unhooking posture of the multi-vehicle adaptive unhooking robot. This approach considers the overall factors of the activity range of the unhooking behavior, the movement path of the multi-vehicle adaptive unhooking robot, and the unhooking posture of the multi-vehicle adaptive unhooking robot, ensuring the accuracy of the adaptive unhooking events. Simultaneously, an unhooking working area is introduced, which, considering the overall factors of each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot, further improves the accuracy of the adaptive unhooking events.

[0073] At this point, the system collects the movement paths of the multi-vehicle adaptive unhooking robot, and generates unhooking behavior events (such as approaching unhooking, aligning unhooking, and executing unhooking) in S132. Each event corresponds to specific triggering conditions and behavior parameters. The system calculates the effective range of activity in space for the behavior based on the type and triggering conditions of the behavior event. The range of activity includes: spatial range: the effective three-dimensional spatial region of the behavior event (such as a spherical or cubic region centered on the unhooking feature); temporal range: the duration of the behavior event (such as within 5 seconds from triggering to completion); dynamic range: the range of activity adjusted according to the robot's speed and posture (such as the range expanding during high-speed movement and shrinking during low-speed movement); range calculation: spatial range is calculated using spatial geometric algorithms (such as convex hull algorithm and bounding box algorithm); temporal range is calculated using time series analysis; and dynamic range is adjusted using dynamic programming algorithms.

[0074] The adaptive unhooking event is determined based on the activity range, movement path, and unhooking posture. Activity range: the spatial, temporal, and dynamic range of each unhooking behavior event; movement path: the robot's actual movement trajectory and planned path; unhooking posture: the robot's arm posture, gripper state, and direction of movement, etc.

[0075] The system determines whether the robot has entered the valid range of a certain behavioral event by comparing the robot's movement path and activity range in real time. When the robot enters the activity range, the system generates an adaptive unhooking event based on the unhooking posture and behavioral parameters. The adaptive unhooking event includes: event type (such as adaptive approach unhooking, adaptive alignment unhooking); triggering condition (such as entering the activity range and the posture meeting the requirements); execution action (such as adjusting the arm angle, reducing the movement speed); anomaly handling (such as stopping the action when exceeding the activity range); event optimization: reinforcement learning algorithm is used to optimize the adaptive unhooking event to improve the success rate and efficiency of the event; multi-objective optimization is introduced to balance unhooking speed, accuracy and safety.

[0076] Specifically, assuming the multi-car coupling robot needs to perform coupling and uncoupling operations on three different types of carriages (Type A / B / C) in a railway freight yard, the following path data is collected: Starting point: Freight yard entrance (X=0, Y=0, Z=0); Target point: Type C carriage location (X=100, Y=50, Z=0); Path nodes: N1: Entrance turning point (X=20, Y=10, Z=0); N2: Near Type A carriage (X=40, Y=20, Z=0); N3: Near Type B carriage (X=60, Y=30, Z=0); N4: Near Type C carriage (X=80, Y=40, Z=0); Speed ​​curve: 0-20m: Accelerate to 1m / s; 20-80m: Constant speed 1m / s; 80-100m: Decelerate to 0.2m / s.

[0077] Behavioral events (taking a C-type carriage as an example): BE7: Approaching to uncoupling (trigger condition: position event E7 action event C1); BE8: Aligning to uncoupling (trigger condition: position event E8 action event C2); BE9: Executing uncoupling (trigger condition: position event E9 action event C3); Activity range calculation: BE7 Activity range: Spatial range: a spherical area with a radius of 5 meters centered on N4; Time range: within 10 seconds from entering N4 to N8; Dynamic range: radius of 5 meters at a speed of 1 m / s, radius of 2 meters at a speed of 0.2 m / s; BE8 Activity range: Spatial range: a spherical area with a radius of 1 meter centered on N8; Time range: within 5 seconds from arriving at N8 to starting uncoupling; Dynamic range: adjusted according to the outrigger posture; BE9 Activity range: Spatial range: a spherical area with a radius of 0.5 meters centered on N9; Time range: within 3 seconds of the uncoupling action; Dynamic range: adjusted according to the gripper status.

[0078] Path and activity range comparison: The robot moves to N4 (X=80, Y=40, Z=0), entering the activity range of BE7; the system detects that the robot speed is 0.5m / s and the posture is that the arm is extended to 30 degrees; Adaptive unhooking event generation: Event AE1: Adaptive approach unhooking trigger condition: entering the activity range of BE7 and speed <1m / s; Action execution: reduce speed to 0.2m / s, extend arm to 45 degrees; Anomaly handling: if it exceeds the BE7 range, stop the action and alarm; The robot continues to move to N8 (X=90, Y=45, Z=0), entering the activity range of BE8; the system detects that the robot speed is 0.2m / s and the posture is that the arm is aligned with the C-type connector.

[0079] Adaptive Unhooking Event Generation: Event AE2: Adaptive Alignment Unhooking; Triggering Condition: Entering the BE8 activity range and the outrigger aligns with the connector; Execution Action: Fine-tuning the outrigger angle to ensure alignment accuracy; Anomaly Handling: If alignment fails, replan the path; The robot moves to N9 (X=95, Y=48, Z=0) and enters the BE9 activity range; The system detects that the robot speed is 0m / s and the posture is gripper closed; Adaptive Unhooking Event Generation: Event AE3: Adaptive Unhooking Execution; Triggering Condition: Entering the BE9 activity range and the gripper is closed; Execution Action: The gripper releases, completing the unhooking; Anomaly Handling: If unhooking fails, increase the clamping force and retry.

[0080] Output: The system generates 3 adaptive unhooking events (AE1-AE3), each event includes: triggering conditions (entering the activity range + posture / speed requirements); execution actions (speed adjustment, outrigger control, gripper operation); and exception handling strategies (stop, alarm, retry).

[0081] refer to Figure 5 In step S14, the specific steps are as follows:

[0082] S141: Collect each hook removal location, and determine the corresponding hook removal environment area based on each hook removal location and corresponding hook removal characteristics. Determine multiple environmental parameters based on real-time detection of the hook removal environment area, and determine multiple environmental features based on the location, parameter type, and morphology of the hook removal environment area.

[0083] S142: Determine the working environment of the multi-vehicle adaptive unhooking robot based on multiple environmental features and the surrounding space of each unhooking position, and mark the working condition type of the multi-vehicle adaptive unhooking robot according to the working environment.

[0084] S143: Determine the first behavior trajectory based on each working condition type and the unhooking position; determine the second behavior trajectory based on each working condition type and the corresponding adaptive unhooking event; and determine the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on the first behavior trajectory, the second behavior trajectory, and the unhooking behavior matching table.

[0085] In the embodiments of this application, each hook removal location is collected, and the corresponding hook removal environment area is determined based on each hook removal location and the corresponding hook removal features. Multiple environmental parameters are determined based on the real-time detection of the hook removal environment area. Multiple environmental features are determined based on the location, parameter type, and shape of the hook removal environment area of ​​the multiple environmental parameters. This approach takes into account the overall consideration of the location, parameter type, and shape of the hook removal environment area of ​​the multiple environmental parameters, ensuring the accuracy of the multiple environmental features.

[0086] At this point, the system collects data on each unhooking position. Based on the spatial coordinates of the unhooking position and the geometric structure of the unhooking features (such as size, shape, and operating space requirements), the system dynamically constructs a three-dimensional spatial region, namely the "unhooking environment region". The shape of the region can be spherical, cubic, cylindrical, etc., depending on the type of unhooking feature and the operating requirements. For example, Type A connectors require a larger operating space, so the region is set as a spherical region with a radius of 1.5 meters; Type C connectors require more precise operation, so the region is set as a cubic region with a length and width of 0.8 meters and a height of 1.2 meters.

[0087] The system uses multi-sensor fusion (such as RGB-D cameras, LiDAR, infrared sensors, and ultrasonic sensors) to perform real-time scanning of the unhooking environment. The detection frequency is typically 10–30 Hz to ensure the real-time performance and accuracy of the data. The system extracts the following environmental parameters from the sensor data: Spatial parameters: obstacle positions (such as other vehicles, equipment, walls); ground flatness (tilt angle calculated from point cloud data); spatial height limitations (such as overhead pipes, ceiling height); Physical parameters: temperature (affecting robotic arm lubrication and sensor performance); humidity (affecting vision system clarity); light intensity (affecting camera image quality); Dynamic parameters: speed and direction of moving objects (such as people walking, vehicles moving); wind speed (affecting robotic arm stability).

[0088] The system generates multiple environmental features based on the spatial distribution, parameter type, and geometric shape of the de-hooked environmental area using a feature extraction algorithm. The feature generation methods include: clustering obstacle locations into "dense areas" or "sparse areas"; calculating the mean and variance of ground flatness and judging it as "flat" or "uneven"; and judging it as "strong light" or "weak light" based on the light intensity threshold (e.g., 500 lux). Typical environmental features include: obstacle density (high / medium / low); ground flatness (flat / uneven); lighting conditions (strong light / weak light / no light); dynamic interference level (high / medium / low); and spatial openness (open / semi-open / closed).

[0089] Specifically, at a railway freight station, a multi-car uncoupling robot needs to perform uncoupling operations on three different types of train carriages: uncoupling position 1: Type A connector, coordinates (10,5,1.2); uncoupling position 2: Type B connector, coordinates (15,8,1.0); uncoupling position 3: Type C connector, coordinates (20,12,1.5).

[0090] The system collected three unhooking locations and defined the environmental areas for each: Type A connector: spherical area, radius 1.5 meters; Type B connector: cylindrical area, radius 1.0 meter, height 1.5 meters; Type C connector: cubic area, length and width 0.8 meters, height 1.2 meters. The system scanned the three areas using LiDAR and a camera, obtaining the following parameters: Area 1 (Type A): Obstacle locations: two toolboxes are located at (9.5, 4.8, 0) and (10.5, 5.2, ...). 0); Ground flatness: tilt angle 2°; Illumination intensity: 800 lux; Dynamic objects: None; Area 2 (Type B): Obstacle location: 1 moving car located at (14.8,7.9,0), speed 0.5m / s; Ground flatness: tilt angle 5°; Illumination intensity: 300 lux; Dynamic objects: moving car; Area 3 (Type C): Obstacle location: None; Ground flatness: tilt angle 1°; Illumination intensity: 100 lux; Dynamic objects: None.

[0091] The system generates environmental features based on parameters: Region 1: Obstacle density: Medium (2 obstacles); Ground flatness: Flat (tilt angle <3°); Lighting conditions: Strong light (>500 lux); Dynamic interference level: Low (no moving objects); Region 2: Obstacle density: Low (1 obstacle); Ground flatness: Uneven (tilt angle >3°); Lighting conditions: Weak light (<500 lux); Dynamic interference level: Medium (slow-moving objects present); Region 3: Obstacle density: Low (no obstacles); Ground flatness: Flat (tilt angle <3°); Lighting conditions: No light (<200 lux); Dynamic interference level: Low (no moving objects); Output results: The system generates the following environmental feature vectors for the three unhooking positions: Position 1: [Medium, Flat, Strong light, Low]; Position 2: [Low, Uneven, Weak light, Medium]; Position 3: [Low, Flat, No light, Low].

[0092] Furthermore, the working environment of the multi-vehicle adaptive unhooking robot is determined based on multiple environmental features and the surrounding space of each unhooking position. The working condition type of the multi-vehicle adaptive unhooking robot is marked according to the working environment. This takes into account the overall consideration of multiple environmental features and the surrounding space of each unhooking position, ensuring the accuracy of the working environment of the multi-vehicle adaptive unhooking robot.

[0093] At this point, multiple environmental features are generated by step S141. Each unhooking position corresponds to a set of environmental feature vectors. For example: Position 1: [Obstacle density: medium, ground flatness: flat, lighting conditions: strong light, dynamic interference level: low]; Position 2: [Obstacle density: low, ground flatness: uneven, lighting conditions: weak light, dynamic interference level: medium]; Position 3: [Obstacle density: low, ground flatness: flat, lighting conditions: no light, dynamic interference level: low].

[0094] The system analyzes the spatial structure surrounding each unhooking location, including: spatial degrees of freedom: whether there is sufficient operating space (e.g., whether it is restricted above or to the sides); obstacle distribution: the location, size, and mobility of obstacles; ground conditions: ground slope, coefficient of friction, and whether there are steps or potholes; dynamic factors: whether there are moving vehicles, personnel, or other equipment; the system integrates environmental characteristics with surrounding spatial information to construct a multi-dimensional description of the working environment; the working environment includes: static environment: fixed elements (e.g., walls, fixed equipment); dynamic environment: changing elements (e.g., moving vehicles, changes in lighting); operational constraints: restrictions on robot movements (e.g., minimum operating distance, maximum permissible speed).

[0095] The system has multiple preset working condition types, each corresponding to specific environmental conditions and operational requirements. Common working condition types include: Standard working condition: good environment, no significant obstacles, sufficient lighting, and flat ground; Complex working condition: obstacles or dynamic interference exist, lighting is weak, and the ground is slightly uneven; Harsh working condition: dense obstacles, extremely poor lighting, severely uneven ground, and high-frequency dynamic interference.

[0096] The system automatically labels the working condition type based on a comprehensive score of the working environment. Example scoring rules: Standard working condition: Obstacle density: Low; Ground smoothness: Smooth; Lighting conditions: Strong light; Dynamic interference level: Low; Complex working condition: Obstacle density: Medium or low; Ground smoothness: Smooth or uneven; Lighting conditions: Weak light; Dynamic interference level: Medium; Severe working condition: Obstacle density: High; Ground smoothness: Uneven; Lighting conditions: No light; Dynamic interference level: High; Labeling process: The system quantifies and scores each environmental feature (e.g., low = 1 point, medium = 2 points, high = 3 points); calculates the total score for the working environment; and maps the total score range to the corresponding working condition type.

[0097] Specifically, the multi-vehicle adaptive unhooking robot needs to perform unhooking tasks at three unhooking positions (P1, P2, P3); the environmental characteristics of each position are as follows: P1: obstacle density: medium; ground flatness: flat; lighting conditions: strong light; dynamic interference level: low; P2: obstacle density: low; ground flatness: uneven; lighting conditions: weak light; dynamic interference level: medium; P3: obstacle density: low; ground flatness: flat; lighting conditions: no light; dynamic interference level: low.

[0098] Operating environment determined as follows: P1: Surrounding space: There are 2 fixed obstacles, the ground is flat, and there is no dynamic interference; Operating environment: Mainly static environment with few operational constraints; P2: Surrounding space: There is 1 movable obstacle, the ground is tilted at 5 degrees, and there are slow-moving vehicles; Operating environment: Significant dynamic environment, operation requires obstacle avoidance and low speed; P3: Surrounding space: No obstacles, the ground is flat, but the lighting is extremely poor; Operating environment: Static environment, but requires reliance on non-visual sensors.

[0099] Operating condition type label: P1: Score: Medium (2) + Smooth (1) + Strong light (1) + Low (1) = 5 points; Operating condition type: Standard operating condition; P2: Score: Low (1) + Unsmooth (2) + Weak light (2) + Medium (2) = 7 points; Operating condition type: Complex operating condition; P3: Score: Low (1) + Smooth (1) + No light (3) + Low (1) = 6 points; Operating condition type: Complex operating condition (Special processing is required due to no light); The system generates the following operating condition type labels for the three unhooking positions: P1: Standard operating condition; P2: Complex operating condition; P3: Complex operating condition; These operating condition types will be used as inputs to step S143 to generate adaptive unhooking behavior.

[0100] Therefore, the first behavior trajectory is determined based on each working condition type and the unhooking position, the second behavior trajectory is determined based on each working condition type and the corresponding adaptive unhooking event, and the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot is determined based on the first behavior trajectory, the second behavior trajectory and the unhooking behavior matching table. This takes into account the overall consideration of the first behavior trajectory, the second behavior trajectory and the unhooking behavior matching table, and ensures the accuracy of the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot.

[0101] At this point, the first action trajectory is determined based on the various working conditions and the hook-off position. Each hook-off position is generated by S141, such as P1, P2, and P3. Each working condition type is marked by S142, such as P1 for "standard working condition" and P2 and P3 for "complex working conditions." The first action trajectory is a preliminary sequence of actions generated based on the static working condition type and hook-off position. It does not consider real-time dynamic events, but only generates the basic path and action logic according to preset rules. Each trajectory point includes: position coordinates (x, y, z); action type (e.g., approach, align, grab, release); speed and direction control parameters. Optionally, the system can call a preset trajectory template based on the spatial coordinates of the hook-off position and the working condition type. For example: standard working condition: uses a high-speed straight trajectory for rapid approach; complex working condition: uses a low-speed curved trajectory for obstacle avoidance priority.

[0102] The second behavior trajectory is determined based on various working condition types and corresponding adaptive unhooking events. Working condition types include "standard working condition" and "complex working condition." Adaptive unhooking events are generated by S133, such as AE1, AE2, and AE3. The second behavior trajectory is a dynamically adjusted trajectory generated based on the first behavior trajectory, combined with real-time event triggers. It includes event-driven action changes, such as pausing or detouring when encountering an obstacle; switching to infrared mode when there is insufficient light; and automatically retrying when unhooking fails. Optionally, the system maps adaptive unhooking events to trajectory nodes, forming "condition-action" branches. Each branch point includes: a trigger condition (e.g., "obstacle detected"); an action response (e.g., "pause and replan the path"); and a recovery strategy (e.g., "detour and continue the original trajectory").

[0103] The adaptive unhooking behavior is determined based on the first behavior trajectory, the second behavior trajectory, and the unhooking behavior matching table. The first behavior trajectory is the basic action sequence; the second behavior trajectory is the dynamically adjusted action sequence; and the unhooking behavior matching table is a predefined behavior rule library containing the optimal behavior under different combinations of working conditions and events. The unhooking behavior matching table is collected, as shown in Table 1.

[0104] Table 1 Unhooking Behavior Matching Table

[0105] Operating conditions Event Type Behavioral suggestions Speed ​​control Torque control Standard operating conditions normal approach Approaching quickly high speed medium Standard operating conditions Unhooking failed Try again medium speed high Complex working conditions Encountering obstacles Obstacle avoidance low speed Low Complex working conditions Insufficient light Switching sensors medium speed medium

[0106] The system fuses the first and second behavioral trajectories to generate a final behavioral sequence; for each trajectory point, it queries the matching table to determine the optimal behavioral parameters; the final behavior includes: action sequence (such as "approach → align → grab → release"); control parameters (speed, torque, sensor mode); and abnormal handling mechanisms (such as retry, alarm, emergency stop).

[0107] Specifically, there are three unhooking positions: P1, P2, and P3; operating conditions: P1: standard operating condition; P2: complex operating condition; P3: complex operating condition; adaptive unhooking events: AE1: P1 approaches normally; AE2: P2 encounters an obstacle; AE3: P3 has insufficient lighting.

[0108] Generate the first line trajectory: P1 (standard working condition): trajectory: straight line fast approach → alignment → grab → release; P2 (complex working condition): trajectory: curve slow approach → alignment → grab → release; P3 (complex working condition): trajectory: curve slow approach → alignment → grab → release.

[0109] Generate the second behavior trajectory: P1 (AE1: normal approach): no adjustment, follow the first trajectory; P2 (AE2: encounter obstacle): adjust to: approach → detect obstacle → detour → approach again → align → grab → release; P3 (AE3: insufficient light): adjust to: approach → switch infrared sensor → align → grab → release.

[0110] Query the hook-off behavior matching table: P1: Matches "Standard working condition + normal approach" → Behavior: fast approach, high speed, medium torque; P2: Matches "Complex working condition + encountering obstacles" → Behavior: detour to avoid obstacles, low speed, low torque; P3: Matches "Complex working condition + insufficient lighting" → Behavior: switch sensors, medium speed, medium torque.

[0111] The system generates three adaptive unhooking behaviors: P1 behavior: Action sequence: rapid approach → alignment → gripping → release; Control parameters: high speed, medium torque; Anomaly handling: none; P2 behavior: Action sequence: low speed approach → detour → reapproach → alignment → gripping → release; Control parameters: low speed, low torque; Anomaly handling: alarm when obstacle avoidance fails; P3 behavior: Action sequence: low speed approach → switch infrared sensor → alignment → gripping → release; Control parameters: medium speed, medium torque; Anomaly handling: retry when sensor switching fails. These adaptive unhooking behaviors will serve as the final control commands, input into the robot execution system, to achieve efficient and safe multi-vehicle adaptive unhooking operations.

[0112] refer to Figure 6 In step S15, the specific steps are as follows:

[0113] S151: In each adaptive unhooking behavior, the working path of the unhooking working arm of the multi-vehicle adaptive unhooking robot is determined according to the adaptive unhooking behavior, and the corresponding unhooking work efficiency is estimated according to the working path of the unhooking working arm, the working rhythm of the unhooking working arm and the motion dimension of the adaptive unhooking behavior.

[0114] S152: Collect multiple unhooking positions, determine the unhooking work level of the multi-vehicle adaptive unhooking robot based on the multiple unhooking positions and the corresponding unhooking work efficiency, and collect the remaining work content of the multi-vehicle adaptive unhooking robot. Based on the identification of the remaining work content of the multi-vehicle adaptive unhooking robot, determine multiple sub-work lists, and determine the corresponding remaining unhooking behavior and corresponding time nodes based on the detection of multiple sub-work lists.

[0115] S153: Determine the remaining unhooking operation system of the multi-vehicle adaptive unhooking robot based on each remaining unhooking behavior, the corresponding time node, and the unhooking operation level of the multi-vehicle adaptive unhooking robot, and realize the adaptive control of the multi-vehicle adaptive unhooking robot on the remaining unhooking position under the trigger of the remaining unhooking operation system.

[0116] In the embodiments of this application, in each adaptive unhooking behavior, the working path of the unhooking working arm of the multi-vehicle adaptive unhooking robot is determined according to the adaptive unhooking behavior. The corresponding unhooking efficiency is estimated based on the working path of the unhooking working arm, the working rhythm of the unhooking working arm and the motion dimension of the adaptive unhooking behavior. This incorporates the overall consideration of the adaptive unhooking behavior and ensures the accuracy of the working path of the unhooking working arm of the multi-vehicle adaptive unhooking robot.

[0117] At this time, each adaptive unhooking behavior (from S143) is as follows: P1: fast approach → align → grab → release; P2: slow approach → detour → approach again → align → grab → release; P3: slow approach → switch infrared sensor → align → grab → release.

[0118] The working path is the trajectory of the end effector (gripper) in three-dimensional space when the outrigger performs the unhooking action. Each action corresponds to a spatial path point, and all path points constitute the complete working path. The path point includes: coordinates (x, y, z); attitude (pitch angle, yaw angle, roll angle); and velocity and acceleration constraints. The system selects different path planning algorithms according to the action type and working condition: straight-line actions (such as rapid approach): linear interpolation algorithm is used to generate a straight-line path; curved actions (such as detour): Bézier curve or spline interpolation is used to generate a smooth curved path; sensor switching actions: sensor switching action points are inserted in the path to ensure that the outrigger position is stable when the sensor switches.

[0119] The work cycle time refers to the time interval between each action. The length of the cycle time directly affects the efficiency of unhooking: short cycle time → high efficiency; long cycle time → low efficiency. The system sets cycle time parameters based on the action type and working condition: standard working condition (e.g., P1): short cycle time (e.g., 1 second / action); complex working condition (e.g., P2, P3): long cycle time (e.g., 2 seconds / action). The cycle time parameters can be dynamically adjusted: if an obstacle is detected, the cycle time automatically increases; if path optimization is successful, the cycle time automatically decreases. The formula for calculating work efficiency is:

[0120]

[0121] Total beat time = sum of all action beats; Action dimension score = Spatial dimension score + Control dimension score.

[0122] Action dimension refers to the complexity of an action, including: spatial dimension (such as straight line, curve, and detour); control dimension (such as speed, torque, and sensor switching); the system scores each action dimension: straight line action: low score (e.g., 1 point); curve action: medium score (e.g., 2 points); detour action: high score (e.g., 3 points); sensor switching action: high score (e.g., 3 points); the higher the score, the more complex the action and the lower the efficiency; the system calculates the efficiency of each unhooking action based on the action dimension score and the total cycle time; for example: P1: total cycle time = 4 seconds, action dimension score = 4 points → efficiency = 1 / 4 × 4 = 1; P2: total cycle time = 12 seconds, action dimension score = 12 points → efficiency = 1 / 12 × 12 = 1; P3: total cycle time = 10 seconds, action dimension score = 10 points → efficiency = 1 / 10 × 10 = 1.

[0123] Furthermore, multiple unhooking locations are collected, and the unhooking work level of the multi-vehicle adaptive unhooking robot is determined based on the multiple unhooking locations and the corresponding unhooking work efficiency. At the same time, the remaining work content of the multi-vehicle adaptive unhooking robot is collected, and multiple sub-work lists are determined based on the identification of the remaining work content of the multi-vehicle adaptive unhooking robot. Based on the detection of multiple sub-work lists, the corresponding remaining unhooking behavior and corresponding time nodes are determined. This overall consideration of the detection of multiple sub-work lists ensures the accuracy of the corresponding remaining unhooking behavior and corresponding time nodes.

[0124] At this time, multiple unhooking positions are collected, such as P1, P2, and P3. Each position includes spatial coordinates and working condition type. The corresponding unhooking efficiency is calculated by S151, for example: P1: efficiency 1.0; P2: efficiency 0.58; P3: efficiency 0.7.

[0125] The work level is a comprehensive evaluation of the difficulty of executing the dehook position, used to guide subsequent task scheduling; the level classification standards are as follows: High efficiency level (level A): efficiency ≥ 0.9, indicating simple action, fast execution, and low resource consumption; Medium efficiency level (level B): 0.6 ≤ efficiency < 0.9, indicating moderate action with some complexity; Low efficiency level (level C): efficiency < 0.6, indicating complex action, slow execution, and high resource consumption; the system automatically maps to the corresponding level based on the work efficiency of the dehook position; for example: P1: efficiency 1.0 → level A; P2: efficiency 0.58 → level C; P3: efficiency 0.7 → level B.

[0126] The system retrieves a list of unfinished unhooking tasks from the task management module. Each task includes: unhooking location (e.g., P4, P5, P6); working condition type (e.g., standard, complex, obstacle-dense); priority (high, medium, low); and estimated completion time. The sub-work list categorizes and groups the remaining tasks for easier scheduling and execution. Categorization criteria include: grouping by location area (e.g., area A, area B); grouping by working condition type (e.g., standard working condition group, complex working condition group); and grouping by priority (e.g., high priority group, medium priority group). The system divides the remaining tasks into multiple sub-work lists according to preset rules.

[0127] The system generates corresponding unhooking behaviors for each task in the sub-work list. The behavior generation method is to search the behavior matching table and match the best behavior template based on the work condition type and location characteristics. For example: P4: Standard work condition → fast approach, alignment, grab and release; P5: Complex work condition → slow approach, detour, re-approach, alignment, grab and release; P6: Obstacle-dense work condition → slow approach, switch infrared sensor, alignment, grab and release.

[0128] The time node is the planned execution time for each decoupling action; the calculation method is: based on the current time, task priority, and estimated execution time, the time node is dynamically scheduled; high-priority tasks are scheduled first, and low-priority tasks are postponed; for example: current time: 10:00:00; P4: high priority, estimated execution time 5 seconds → time node 10:00:00; P5: medium priority, estimated execution time 10 seconds → time node 10:00:10; P6: low priority, estimated execution time 8 seconds → time node 10:00:25.

[0129] Therefore, based on each remaining unhooking action, the corresponding time node, and the unhooking work level of the multi-vehicle adapted unhooking robot, the remaining unhooking work system of the multi-vehicle adapted unhooking robot is determined. Under the triggering of the remaining unhooking work system, the multi-vehicle adapted unhooking robot achieves adaptive control of the remaining unhooking positions. This system takes into account the overall consideration of each remaining unhooking action, the corresponding time node, and the unhooking work level of the multi-vehicle adapted unhooking robot, ensuring the accuracy of the remaining unhooking work system of the multi-vehicle adapted unhooking robot. At the same time, the adaptive unhooking behavior of the multi-vehicle adapted unhooking robot is introduced, realizing the overall consideration of unhooking work efficiency, corresponding unhooking positions, and the remaining work content of the multi-vehicle adapted unhooking robot, thus improving the accuracy of the remaining unhooking work system.

[0130] At this point, the remaining unhooking actions are generated by step S152, for example: P4: fast approach → alignment → grasp → release; P5: slow approach → detour → re-approach → alignment → grasp → release; P6: slow approach → switch infrared sensor → alignment → grasp → release; corresponding time nodes: for example: P4: 10:00:00; P5: 10:00:10; P6: 10:00:25; unhooking work level: determined by S152, for example: P4: Level A (high efficiency); P5: Level B (medium efficiency); P6: Level C (low efficiency); The remaining unhooking work system is the overall scheduling and execution framework for the robot's remaining unhooking tasks; it includes the following elements: behavior sequence: the specific action sequence for each unhooking position; time node: the start time of each behavior; work level: used to dynamically adjust resource allocation and priority; exception handling: different exception response strategies are set for tasks of different levels.

[0131] Work system construction method: The system sorts tasks according to time nodes and generates a task execution queue; resources are dynamically adjusted according to work level: A-level tasks: high-speed channels are prioritized to reduce waiting time; B-level tasks: standard resources are used with appropriate delays; C-level tasks: low-speed channels are used, and can be preempted by high-priority tasks; exception handling strategy: A-level tasks: retry immediately after failure; B-level tasks: retry with a delay after failure; C-level tasks: log after failure and do not retry.

[0132] When the system reaches the first time node (e.g., 10:00:00), it automatically starts the work system. Triggering conditions: time node reached; robot idle; no abnormalities in the working environment. Execution flow: The system extracts the task for the current time node (e.g., P4) from the work system; allocates resources according to the task's work level (Level A); executes the unhooking behavior sequence: rapid approach → alignment → grab → release; upon completion, the system checks the next time node (10:00:10) and starts P5; this process continues until all tasks are completed. Dynamic adjustment: If P4 times out, the system automatically shifts the time nodes of P5 and P6 to the next period; if P6 is preempted by a higher-priority task, the system places it in a waiting queue and reschedules it when resources are available; exception handling: If P5 fails, the system retry after a 5-second delay according to the Level B strategy; if P6 fails, the system logs and skips the task.

[0133] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of the multi-vehicle adaptive unhooking robot working condition adaptive adjustment system in an embodiment of the present invention; the multi-vehicle adaptive unhooking robot working condition adaptive adjustment system includes:

[0134] The area working image module 21 is used to determine multiple area nodes based on the movement path of the multi-vehicle adaptive unhooking robot, and to acquire multiple area working images based on the detection of multiple area nodes.

[0135] The multimodal data module 22 is used to determine multiple environmental features, dynamic features, and de-hooking features based on image recognition of working images in various regions, and to determine the de-hooking working area based on the multiple environmental features, dynamic features, and de-hooking features.

[0136] The adaptive unhooking event module 23 is used to determine the unhooking behavior event of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot, and to determine the adaptive unhooking event of the multi-vehicle adaptive unhooking robot based on each unhooking behavior event and the movement path of the multi-vehicle adaptive unhooking robot.

[0137] The adaptive unhooking behavior module 24 is used to determine the working condition type of the multi-vehicle adaptive unhooking robot based on each unhooking position and multiple environmental features, and to determine the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each working condition type and the corresponding adaptive unhooking event.

[0138] The remaining unhooking system module 25 is used to predict the corresponding unhooking efficiency based on each adaptive unhooking behavior, and to determine the remaining unhooking system of the multi-vehicle adaptive unhooking robot based on the unhooking efficiency, the corresponding unhooking position and the remaining work content of the multi-vehicle adaptive unhooking robot.

[0139] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A multi-vehicle adaptive picking hook robot working condition self-adaptive adjustment method, characterized in that, include: Multiple regional nodes are determined based on the movement path of the multi-vehicle adaptive unhooking robot, and multiple regional working images are collected based on the detection of multiple regional nodes. Based on image recognition of working images in various regions, multiple environmental features, dynamic features, and de-hooking features are determined, and the de-hooking working area is determined based on these features. Based on the unhooking work area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adapted unhooking robot, the unhooking behavior events of the multi-vehicle adapted unhooking robot relative to the unhooking feature are determined. Based on each unhooking behavior event and the movement path of the multi-vehicle adapted unhooking robot, adaptive unhooking events of the multi-vehicle adapted unhooking robot are determined, including: acquiring the unhooking work area; determining multiple unhooking working nodes based on the identification of the unhooking work area; determining position events based on the locations of the multiple unhooking working nodes and the location of the unhooking feature; and determining the position events based on the multiple unhooking working nodes and the multi-vehicle adapted... The unhooking robot's unhooking working arm determines motion events. Based on position events, motion events, and the working mode of the multi-vehicle adapted unhooking robot relative to the unhooking working area, the unhooking behavior events of the multi-vehicle adapted unhooking robot relative to the unhooking feature are determined. The movement path of the multi-vehicle adapted unhooking robot is collected, and the activity range of the corresponding unhooking behavior is determined according to the detection of each unhooking behavior event. Based on the activity range of the unhooking behavior, the movement path of the multi-vehicle adapted unhooking robot, and the unhooking posture of the multi-vehicle adapted unhooking robot, the adaptive unhooking event of the multi-vehicle adapted unhooking robot is determined. The working conditions of the multi-vehicle adaptive unhooking robot are determined based on each unhooking location and multiple environmental features. The adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot is determined based on each working condition and the corresponding adaptive unhooking event. Based on the estimated unhooking efficiency of each adaptive unhooking behavior, the remaining unhooking system of the multi-vehicle adaptive unhooking robot is determined according to the unhooking efficiency, the corresponding unhooking position, and the remaining work of the multi-vehicle adaptive unhooking robot.

2. The multi-vehicle adaptive picking hook robot working condition self-adaptive adjustment method according to claim 1, characterized in that, The process of determining multiple regional nodes based on the movement path of the multi-vehicle adaptive unhooking robot, and acquiring multiple regional working images based on the detection of these multiple regional nodes, includes: The current and target positions of the multi-vehicle adaptive unhooking robot are collected. Based on the current position, target position and corresponding movement space, the movement path of the multi-vehicle adaptive unhooking robot is determined. Based on the detection of the movement path of the multi-vehicle adaptive unhooking robot, multiple unhooking operation areas are determined, and the area nodes of each unhooking operation area are marked. In each regional node, multiple cameras are triggered based on the node's location. These cameras then take a circular view of the unhooking operation area where the regional node is located and collect images of multiple sub-regions. Based on these sub-region images, the corresponding unhooking locations, and the previous positioning locations of the multi-vehicle adaptive unhooking robots, the corresponding regional working image is determined.

3. The multi-vehicle adaptive unhooking robot working condition self-adaptive adjustment method according to claim 1, characterized in that, The process involves determining multiple environmental features, dynamic features, and de-hooking features based on image recognition of working images in various regions, and determining the de-hooking working area based on these features, including: In each region working image, multiple sub-feature images are determined based on the recognition of the region working image. The corresponding image recognition mode is determined based on the position, image contour and image type weight of the multiple sub-feature images. Different types of sub-feature images are recognized using the corresponding image recognition mode. The corresponding image recognition mode is executed on multiple sub-feature images, and the corresponding feature combination is output. The region working image has multiple feature combinations. Multiple environmental features, dynamic features, and de-hooking features are determined by screening multiple feature combinations. A preliminary region is constructed based on the multiple environmental features and dynamic features. The de-hooking work area is determined based on the preliminary region and the de-hooking features.

4. The multi-vehicle adaptive unhooking robot working condition self-adaptive adjustment method according to claim 1, characterized in that, The process involves determining the operating condition type of the multi-vehicle adaptive unhooking robot based on each unhooking location and multiple environmental features, and determining the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each operating condition type and the corresponding adaptive unhooking event, including: Collect data at each hook removal location, determine the corresponding hook removal environment area based on each hook removal location and corresponding hook removal characteristics, determine multiple environmental parameters based on real-time detection of the hook removal environment area, and determine multiple environmental features based on the location, type, and morphology of the hook removal environment area. The working environment of the multi-vehicle adaptive unhooking robot is determined based on multiple environmental characteristics and the surrounding space of each unhooking position, and the working condition type of the multi-vehicle adaptive unhooking robot is marked according to the working environment.

5. The multi-vehicle adaptive unhooking robot working condition self-adaptive adjustment method according to claim 4, characterized in that, The process of determining the working condition type of the multi-vehicle adaptive unhooking robot based on each unhooking position and multiple environmental features, and determining the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each working condition type and the corresponding adaptive unhooking event, further includes: The first behavior trajectory is determined based on each working condition type and the unhooking position. The second behavior trajectory is determined based on each working condition type and the corresponding adaptive unhooking event. The adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot is determined based on the first behavior trajectory, the second behavior trajectory and the unhooking behavior matching table.

6. The multi-vehicle adaptive unhooking robot working condition self-adaptive adjustment method according to claim 1, characterized in that, The remaining unhooking process of the multi-vehicle adaptive unhooking robot is determined based on the estimated unhooking efficiency corresponding to each adaptive unhooking behavior, the unhooking efficiency, the corresponding unhooking position, and the remaining tasks of the multi-vehicle adaptive unhooking robot. This includes: In each adaptive unhooking behavior, the working path of the unhooking working arm of the multi-vehicle adaptive unhooking robot is determined based on the adaptive unhooking behavior. The corresponding unhooking efficiency is estimated based on the working path of the unhooking working arm, the working rhythm of the unhooking working arm, and the motion dimension of the adaptive unhooking behavior.

7. The multi-vehicle adaptive unhooking robot working condition self-adaptive adjustment method according to claim 6, characterized in that, The method of estimating the decoupling efficiency based on each adaptive decoupling behavior, and determining the remaining decoupling system of the multi-vehicle adaptive decoupling robot based on the decoupling efficiency, the corresponding decoupling position, and the remaining tasks of the multi-vehicle adaptive decoupling robot, further includes: Multiple unhooking locations are collected, and the unhooking work level of the multi-vehicle adaptive unhooking robot is determined based on the multiple unhooking locations and the corresponding unhooking work efficiency. At the same time, the remaining work content of the multi-vehicle adaptive unhooking robot is collected, and multiple sub-work lists are determined based on the identification of the remaining work content of the multi-vehicle adaptive unhooking robot. Based on the detection of multiple sub-work lists, the corresponding remaining unhooking behavior and corresponding time nodes are determined. Based on each remaining unhooking action, the corresponding time node, and the unhooking work level of the multi-vehicle adaptive unhooking robot, the remaining unhooking work system of the multi-vehicle adaptive unhooking robot is determined, and the adaptive control of the remaining unhooking position of the multi-vehicle adaptive unhooking robot is realized under the trigger of the remaining unhooking work system.

8. A multi-vehicle adaptive picking hook robot working condition self-adaptive adjustment system, characterized in that, The multi-vehicle adaptive unhooking robot working condition adaptive adjustment system is applied to the multi-vehicle adaptive unhooking robot working condition adaptive adjustment method as described in any one of claims 1-7, and the multi-vehicle adaptive unhooking robot working condition adaptive adjustment system includes: The regional working image module is used to determine multiple regional nodes based on the movement path of the multi-vehicle adaptive unhooking robot, and to acquire multiple regional working images based on the detection of multiple regional nodes. The unhooking work area module is used to determine multiple environmental features, dynamic features, and unhooking features based on image recognition of each region's working image, and to determine the unhooking work area based on these multiple environmental features, dynamic features, and unhooking features. The adaptive unhooking event module is used to determine the unhooking behavior events of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the unhooking working area, the location of the unhooking feature, and the unhooking working arm of the multi-vehicle adaptive unhooking robot. It then determines the adaptive unhooking events of the multi-vehicle adaptive unhooking robot based on each unhooking behavior event and the robot's movement path. This includes: acquiring the unhooking working area; determining multiple unhooking working nodes based on the identification of the unhooking working area; determining position events based on the locations of the multiple unhooking working nodes and the location of the unhooking feature; and determining the position events based on the multiple unhooking working nodes. The system determines the motion events of the unhooking working arm of the multi-vehicle adaptive unhooking robot, and determines the unhooking behavior events of the multi-vehicle adaptive unhooking robot relative to the unhooking feature based on the position events, motion events, and the working mode of the multi-vehicle adaptive unhooking robot relative to the unhooking working area. The system collects the movement path of the multi-vehicle adaptive unhooking robot, determines the activity range of the corresponding unhooking behavior based on the detection of each unhooking behavior event, and determines the adaptive unhooking event of the multi-vehicle adaptive unhooking robot based on the activity range of the unhooking behavior, the movement path of the multi-vehicle adaptive unhooking robot, and the unhooking posture of the multi-vehicle adaptive unhooking robot. The adaptive unhooking behavior module is used to determine the working condition type of the multi-vehicle adaptive unhooking robot based on each unhooking position and multiple environmental features, and to determine the adaptive unhooking behavior of the multi-vehicle adaptive unhooking robot based on each working condition type and the corresponding adaptive unhooking event. The remaining unhooking system module is used to predict the corresponding unhooking efficiency based on each adaptive unhooking behavior, and to determine the remaining unhooking system of the multi-vehicle adaptive unhooking robot based on the unhooking efficiency, the corresponding unhooking position, and the remaining work content of the multi-vehicle adaptive unhooking robot.

Images