Dynamic unhooking control method and system for unhooking robot based on feedback mechanism

Abstract

The invention discloses an unhooking robot dynamic unhooking control method and system based on a feedback mechanism, and relates to the technical field of robot control, and the method comprises the steps: obtaining an unhooking robot, moving the unhooking robot to a preset unhooking position, and starting a visual sensor to collect unhooking image information and environment perception image information; edge detection and feature extraction are carried out on the unhooking image information, and hook space coordinate information is determined; performing feature abstraction on the environment perception image information to obtain a target unhooking path; and dynamic characteristic parameters and motion constraint information of the unhooking robot are obtained, unhooking analysis is carried out, unhooking control parameters are obtained, and dynamic unhooking control is carried out. The technical problems that in the prior art, an unhooking robot is low in unhooking efficiency and insufficient in control precision, and certain safety risks exist are solved, and the technical effects that the unhooking efficiency and the control precision of the unhooking robot are improved, and operation safety is guaranteed are achieved.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a dynamic unhooking control method and system of an unhooking robot based on a feedback mechanism. Background Art

[0002] In the field of thermal power, coal is widely used in thermal power plants as the main fuel. The coal transportation process involves a large number of hooking equipment and a complex working environment. Traditional manual hooking and unhooking operations have problems such as low efficiency, imprecise control and high safety risks. In order to ensure the stable supply and transportation safety of coal, power plants need to rely on unhooking robots to realize automated operations, thereby improving overall operating efficiency. During the coal transportation process, the layout and state of the hooking equipment are complex and changeable, and the unhooking robot needs to perform high-precision unhooking tasks under various uncertain conditions. Traditional unhooking robots often rely on fixed path planning and static control methods, which are difficult to cope with real-time changes in hooking postures, position deviations and obstacles in the transportation environment. This not only affects the efficiency of unhooking, but may also cause collisions between the robot and the equipment, increasing safety hazards.

[0003] The existing technology has technical problems such as low efficiency and insufficient control accuracy of the hook removing robot, which poses certain safety risks. Summary of the invention

[0004] The present application provides a dynamic hook-removing control method and system for a hook-removing robot based on a feedback mechanism, which is used to solve the technical problems in the prior art of low hook-removing efficiency and insufficient control accuracy of the hook-removing robot, as well as certain safety risks.

[0005] In view of the above problems, the present application provides a dynamic unhooking control method and system for an unhooking robot based on a feedback mechanism.

[0006] In a first aspect of the present application, a dynamic hook removal control method of a hook removal robot based on a feedback mechanism is provided, the method comprising:

[0007] A hook removal robot is obtained, wherein the hook removal robot integrates a visual sensor and an image processing unit, the hook removal robot is moved to a preset hook removal position, and the visual sensor is started to collect and obtain hook removal image information and environmental perception image information; edge detection and feature extraction are performed on the hook removal image information based on the image processing unit to obtain hook position information and hook posture information, and the hook position information is converted into coordinates according to the preset hook removal position to determine the hook space coordinate information; feature abstraction is performed on the environmental perception image information to generate a hook removal environment area map, and path planning is performed on the hook space coordinate information based on the hook removal environment area map to obtain a target hook removal path; power characteristic parameters and motion constraint information of the hook removal robot are obtained, and based on the power characteristic parameters and motion constraint information, hook removal analysis is performed on the hook posture information to obtain hook removal control parameters; dynamic hook removal control is performed on the hook removal robot based on the target hook removal path and the hook removal control parameters, and a feedback mechanism is introduced to perform real-time deviation analysis and compensatory hook removal control on the hook removal control process.

[0008] The second aspect of the present application provides a dynamic hook-removing control system of a hook-removing robot based on a feedback mechanism, the system comprising:

[0009] An image information acquisition module, the image information acquisition module is used to acquire a hook-removing robot, the hook-removing robot integrates a visual sensor and an image processing unit, moves the hook-removing robot to a preset hook-removing position, and simultaneously starts the visual sensor to collect and obtain hook-removing image information and environment perception image information; a hook space coordinate information determination module, the hook space coordinate information determination module performs edge detection and feature extraction on the hook-removing image information based on the image processing unit, obtains hook position information and hook posture information, and performs coordinate conversion on the hook position information according to the preset hook-removing position to determine the hook space coordinate information; a target hook-removing path acquisition module, the target hook-removing path acquisition module is used to determine the hook space coordinate information of the environment perception image. The invention relates to a method for controlling the hook removal process and a method for controlling the hook removal process. The method comprises the following steps: performing feature abstraction on the environment perception image information, generating a hook removal environment area map, performing path planning on the hook spatial coordinate information based on the hook removal environment area map, and obtaining a target hook removal path; a hook removal control parameter acquisition module, wherein the hook removal control parameter acquisition module is used to obtain the power characteristic parameters and motion constraint information of the hook removal robot, and performing hook removal analysis on the hook posture information based on the power characteristic parameters and motion constraint information to obtain hook removal control parameters; and a dynamic hook removal control module, wherein the dynamic hook removal control module performs dynamic hook removal control on the hook removal robot based on the target hook removal path and the hook removal control parameters, and introduces a feedback mechanism to perform real-time deviation analysis and compensatory hook removal control on the hook removal control process.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] Obtain a hook-removing robot, move the hook-removing robot to a preset hook-removing position, and start the visual sensor to collect and obtain hook-removing image information and environmental perception image information; perform edge detection and feature extraction on the hook-removing image information based on the image processing unit to obtain hook position information and hook posture information, and determine the hook space coordinate information according to the preset hook-removing position; perform feature abstraction on the environmental perception image information, generate a hook-removing environmental area map, perform path planning, and obtain the target hook-removing path; obtain the dynamic characteristic parameters and motion constraint information of the hook-removing robot, perform hook-removing analysis on the hook posture information, and obtain the hook-removing control parameters; perform dynamic hook-removing control on the hook-removing robot, and introduce a feedback mechanism to perform real-time deviation analysis and compensatory hook-removing control on the hook-removing control process. The technical effect of improving the hook-removing efficiency and control accuracy of the hook-removing robot and ensuring operational safety is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A schematic flow chart of a dynamic hook-removing control method of a hook-removing robot based on a feedback mechanism provided in an embodiment of the present application;

[0014] Figure 2 A schematic structural diagram of a dynamic hook-removing control system of a hook-removing robot based on a feedback mechanism provided in an embodiment of the present application.

[0015] Explanation of reference numerals: 10 image information acquisition module, 20 hook space coordinate information determination module, 30 target hook removal path acquisition module, 40 hook removal control parameter acquisition module, 50 dynamic hook removal control module. DETAILED DESCRIPTION

[0016] The present application provides a dynamic hook-removing control method and system for a hook-removing robot based on a feedback mechanism, which is used to solve the technical problems in the prior art of low hook-removing efficiency and insufficient control accuracy of the hook-removing robot, as well as certain safety risks.

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] Embodiment 1, as Figure 1 As shown, the present application provides a dynamic hook-removing control method of a hook-removing robot based on a feedback mechanism, the method comprising:

[0019] Step S100: Acquire a hook-removing robot, which integrates a visual sensor and an image processing unit, move the hook-removing robot to a preset hook-removing position, and simultaneously activate the visual sensor to collect and obtain hook-removing image information and environmental perception image information.

[0020] Specifically, first obtain a robot for hook removal, which integrates a visual sensor and an image processing unit. The visual sensor can capture images of the surrounding environment, while the image processing unit is responsible for in-depth analysis and processing of the images collected by the visual sensor to extract key information. Then, move the hook removal robot to the preset hook removal position, which is the initial hook removal point preset according to the actual operation situation before starting the hook removal operation. It has several important characteristics: on the one hand, this position is known, and its specific spatial coordinates and relative position relationship provide a clear reference for subsequent operations. On the other hand, it is relatively fixed and will not change easily within a certain operation cycle, providing a stable basis for the hook removal operation. Moreover, this position can ensure that the robot can stably capture the full picture of the hook, whether it is the shape, size, position or posture of the hook, it can be clearly captured, which is crucial for the subsequent hook removal operation, enabling the robot to better plan and execute the hook removal action. After the robot accurately reaches the preset unhooking position, the visual sensor is started at the same time. At this time, the visual sensor begins to work efficiently, collecting unhooking image information and environmental perception image information respectively. The unhooking image information mainly focuses on the hook itself, providing key data for the subsequent determination of the specific position and posture of the hook, while the environmental perception image information covers the entire working environment around the robot, including existing obstacles, other equipment, and spatial layout, etc., providing an important basis for the robot to fully understand the working environment and formulate reasonable unhooking strategies.

[0021] Step S200: performing edge detection and feature extraction on the hook removal image information based on the image processing unit to obtain hook position information and hook posture information, and performing coordinate conversion on the hook position information according to the preset hook removal position to determine the hook space coordinate information.

[0022] Specifically, in this process, the image processing unit plays a key role. It first performs edge detection on the hook removal image information, identifies the edges of objects in the image, and highlights the contour edges of the hook. Then, it performs feature extraction to extract the unique features of the hook from the image, including shape, color, texture, etc. Through in-depth analysis of these features, the position information of the hook in the image can be determined, that is, the specific coordinates of the hook on the two-dimensional image plane. At the same time, based on the shape, direction and other features of the hook, the posture information of the hook can be further determined, such as the angle and tilt state of the hook. Subsequently, coordinate conversion is performed based on the preset hook removal position, and the preset hook removal position is used as the origin of the spatial coordinates to construct a three-dimensional space coordinate system. In this coordinate system, the hook position information obtained from the hook removal image information is converted, and the hook position coordinates on the two-dimensional image plane are converted into coordinates in three-dimensional space. In this way, the specific position of the hook in the actual space is accurately known, and the spatial coordinate information of the hook is accurately determined for the subsequent hook removal operation, providing a crucial position basis for achieving accurate hook removal operation.

[0023] Step S300: abstracting features of the environmental perception image information to generate a hook-removing environmental area map, and planning a path for the hook space coordinate information based on the hook-removing environmental area map to obtain a target hook-removing path.

[0024] Specifically, the environmental perception image information is first processed, and the features of the environmental perception image information are abstracted to extract key feature information from the complex image. These features include the shape, size, position, color, etc. of objects in the environment, as well as information such as spatial layout. These abstract features are used to generate a hook-off environment area map. This map is an abstract representation of the robot's surrounding environment, which clearly shows the various objects and spatial relationships in the robot's environment. With the hook-off environment area map, the path planning is performed based on this map for the hook spatial coordinate information. Considering various factors in the environment, such as the location of obstacles, the passable area, etc., combined with the spatial coordinates of the hook, the path planning algorithm is used to calculate the best path from the robot's current position to the hook position. In the planning process, the length, safety, efficiency and other factors of the path should be comprehensively considered to ensure that the robot can reach the hook position in the fastest and safest way, and finally obtain the target hook-off path. This path will guide the robot to accurately move to the hook position for hook-off operation in actual operation.

[0025] Step S400: acquiring the power characteristic parameters and motion constraint information of the unhooking robot, and performing unhooking analysis on the hook posture information based on the power characteristic parameters and motion constraint information to obtain unhooking control parameters.

[0026] Specifically, the dynamic characteristic parameters and motion constraint information of the hook-removing robot are first obtained. The dynamic characteristic parameters include the robot's power output capacity, acceleration, speed, etc. These parameters determine the robot's power performance and motion ability when performing the hook-removing operation. The motion constraint information includes the robot's joint range of motion, maximum load, motion accuracy and other limiting conditions to ensure that the robot operates within a safe and feasible range. Based on these dynamic characteristic parameters and motion constraint information, the hook posture information is unhooked and parsed. The hook posture information reflects the position, angle and direction of the hook in space. By analyzing the posture of the hook and the power and motion ability of the robot, the specific control parameters required by the robot when performing the hook-removing operation are determined. These control parameters include the robot's joint angle adjustment, motion speed setting, grasping force control, etc. Based on the dynamic characteristic parameters, it is judged whether the robot has sufficient power to complete the specific hook removal action. At the same time, the motion constraint information is considered to ensure that the operation is within a safe range. Combined with the posture information of the hook, the specific parameter values ​​of each action required for the robot to successfully remove the hook are accurately calculated. The hook removal control parameters obtained in this way will guide the specific actions of the robot when actually performing the hook removal task, ensuring that the robot can complete the hook removal operation accurately and efficiently, while ensuring the safety and stability of the operation.

[0027] Step S500: Dynamically control the unhooking robot based on the target unhooking path and the unhooking control parameters, and introduce a feedback mechanism to perform real-time deviation analysis and compensatory unhooking control on the unhooking control process.

[0028] Specifically, the unhooking robot is firstly dynamically controlled according to the previously obtained target unhooking path and unhooking control parameters, and the target unhooking path is used to provide action guidance for the robot, so that the robot can accurately move toward the hook position. At the same time, the specific actions of the robot are adjusted according to the unhooking control parameters, including the movement angle, speed and grasping force of the joints, so as to ensure that the robot can perform the unhooking operation in the best way. In this process, a feedback mechanism is introduced to obtain various feedback parameters such as the actual position, posture, movement speed, etc. of the robot by real-time monitoring of the unhooking control process. These actual parameters are compared with the preset target state to calculate the deviation between the two. For example, the distance deviation between the actual position and the target position, the angle deviation between the actual posture and the target posture, etc. Next, the deviation is analyzed to determine the direction and size of the deviation, and then the cause of the deviation is analyzed. According to the results of the deviation analysis, a control parameter fine-tuning strategy is formulated. If the actual position deviates greatly from the target position, the movement speed and direction of the robot need to be adjusted; if the actual posture deviates greatly from the target posture, the angle of the joint needs to be adjusted. Finally, based on the control parameter fine-tuning strategy, the dehooking control parameters are fine-tuned to compensate for deviations, and the robot's movements are adjusted in real time, so that the robot can continuously correct deviations and move closer to the target state. This ensures the accuracy and stability of the dehooking process, and ensures that the robot can complete the dehooking task efficiently even in the face of various uncertainties and interferences.

[0029] In a possible implementation, step S200 further includes:

[0030] Step S210: converting the hook removal image information into a hook removal grayscale image, and filtering, denoising and image enhancement are performed on the hook removal grayscale image to obtain a hook removal standard grayscale image.

[0031] Step S220: Perform edge detection on the hook removal standard grayscale image using a Canny edge detector to obtain hook removal edge detection information.

[0032] Step S230: preset a low edge threshold and a high edge threshold, and based on the low edge threshold and the high edge threshold, perform edge contour screening on the hook removal edge detection information to obtain hook removal contour information.

[0033] Step S240: Anchor frame positioning is performed on the hook removal contour information to obtain the hook removal anchor frame information, and the center point is calculated based on the hook removal anchor frame information to obtain the hook position information.

[0034] Step S250: extracting hook posture features based on the hook removal contour information and the hook position information to obtain the hook posture information.

[0035] Specifically, the hook removal image information is first converted into a hook removal grayscale image. This step converts the color image into a grayscale image, reduces the color information of the image, and makes the subsequent processing more focused on the brightness change of the image. Then, the hook removal grayscale image is filtered and denoised, and the median filter algorithm is used to remove the noise interference in the image and improve the image quality. Then, the image enhancement operation is performed, and the histogram equalization method is used to enhance the contrast and clarity of the image, thereby obtaining a hook removal standard grayscale image, which provides a better basis for subsequent edge detection.

[0036] The Canny edge detector is used to perform edge detection on the standard grayscale image of hook removal. The Canny edge detector can effectively detect the edge information in the image. By calculating the gradient amplitude and direction of the image, the edge strength and direction of the pixels in the image are determined, thereby obtaining the hook removal edge detection information. These edge detection information contains important features such as the contours and boundaries of objects in the image.

[0037] First, the edge low threshold and edge high threshold are preset. The setting of these two thresholds is to effectively screen the hook removal edge detection information. The edge low threshold is used to determine the possible edge starting point, while the edge high threshold is used to determine the more obvious and certain edge. For the hook removal edge detection information, it contains multiple different contours. In order to accurately find the contour representing the hook, it is necessary to screen according to the size, shape and other characteristics of the contour. If the size and shape of a contour do not match the expected characteristics of the hook, then it is not the hook contour we are looking for. At the same time, the grayscale of the hook removal edge detection information that is less than the edge low threshold is set to 0, which means that those parts with low grayscale values ​​and not edges are regarded as non-edge areas, and some weaker edge interference is removed. The grayscale greater than the edge high threshold is set to 1, indicating that these parts are more obvious edge areas. Through such processing, the edge information can be more clearly highlighted, which is convenient for the subsequent accurate identification and extraction of the hook contour. Finally, after screening and processing, the hook removal contour information is obtained, which provides key basic data for determining the hook position and posture.

[0038] The hook contour information is used to locate the anchor frame. Anchor frame positioning is a target detection technology that places rectangular frames (anchor frames) of different sizes and proportions in the image to determine the area containing the target object (hook). According to the characteristics of the hook contour information, such as shape and size, the image will be searched for the anchor frame position that is most likely to contain the hook, thereby obtaining the hook anchor frame information. Next, the center point is calculated based on the hook anchor frame information. The geometric center of the anchor frame can be calculated. For example, for a rectangular anchor frame, the midpoint coordinates of its length and width are calculated. The center point coordinates represent the position information of the hook in the image. In this way, the specific position of the hook in the image can be determined more accurately, providing key position data for subsequent hook removal operations.

[0039] The hook posture feature extraction is performed using the hook removal contour information and the hook position information to obtain the hook posture information. Combined with the hook removal contour information, the shape characteristics of the hook are analyzed. For example, by analyzing the curvature of the contour, the degree of curvature of the hook can be determined; observing the symmetry of the contour can determine whether the hook is tilted, etc. At the same time, according to the direction of the contour, the approximate direction of the hook can be preliminarily determined. Then, using the hook position information, the relative position relationship of the hook in space is further determined. For example, by comparing the position of the surrounding objects, the height and horizontal angle of the hook can be determined. Combining this information, the posture characteristics of the hook can be extracted, including the angle (such as pitch angle, yaw angle, roll angle, etc.), direction and tilt state of the hook, and finally the hook posture information is obtained. This posture information is crucial for the hook removal robot to accurately perform the hook removal operation. It can help the robot better plan the action path and adjust the grasping angle to ensure that the hook removal task is completed efficiently and accurately.

[0040] In a possible implementation, step S300 further includes:

[0041] Step S310: performing multi-dimensional feature extraction and feature dimension reduction processing on the environmental perception image information to obtain a multi-dimensional feature set of the dehooking environment.

[0042] Step S320: Perform distributed entity recognition based on the unhooking environment multi-dimensional feature set to obtain an unhooking environment entity set, perform attribute marking on the unhooking environment entity set, and obtain an unhooking environment entity attribute information set.

[0043] Step S330: abstracting the unhooking environment entity set into a map node set, abstracting the edge attributes of the map node set based on the unhooking environment entity attribute information set, and obtaining a map connection edge set.

[0044] Step S340: Performing map topological connection on the map node set based on the map connection edge set to generate the dehooking environment area map.

[0045] Specifically, for environmental perception image information, multi-dimensional feature extraction is first performed. For texture feature extraction, the gray level co-occurrence matrix (GLCM) algorithm is used. This algorithm calculates the gray level co-occurrence relationship of pixel pairs in the image at different directions and distances to obtain a matrix reflecting the characteristics of texture roughness, directionality, etc., from which texture features such as energy, contrast, and correlation can be extracted. For color feature extraction, the color histogram method is used to divide the color space of the image into multiple intervals, and the number of pixels in each interval is counted to describe the color distribution of the image. For edge feature extraction, the Canny edge detection algorithm is used, which can effectively detect the edge contours of objects in the image and provide the shape and boundary information of the objects. Through these methods, features of multiple dimensions such as texture, color, and edge are extracted from the environmental perception image to form a rich initial feature set. Then, feature dimensionality reduction processing is performed, and the principal component analysis (PCA) algorithm is used. The PCA algorithm finds the main components in the data, projects the high-dimensional feature vectors into a low-dimensional space, and removes redundant features while retaining the main feature information. By calculating the covariance matrix of the features, determining the principal component directions of the features, and then projecting the original feature vectors onto these principal component directions, a low-dimensional feature representation is obtained. This not only reduces the redundancy of features, reduces the amount and complexity of subsequent processing, but also improves the effectiveness of the features, allowing the extracted features to more accurately reflect the essential characteristics of the environment, and ultimately obtaining a multi-dimensional feature set of the dehooking environment.

[0046] Distributed entity recognition is performed based on the multi-dimensional feature set of the dehooking environment. The target detection algorithm in deep learning is used to identify various entities in the environment. The location and category of different objects in the image are determined by analyzing the multi-dimensional features, thereby obtaining the dehooking environment entity set. Then, this entity set is attributed and specific attribute information is given to each identified entity, such as size, shape, and passability. In this way, a set of attribute information of the dehooking environment entities is obtained, which provides more detailed information for subsequent map construction.

[0047] Abstracting the set of unhooking environment entities into a set of map nodes means converting various entities in the actual environment, such as roads, obstacles, road signs, etc., into abstract map nodes. Each entity corresponds to a node, and a node set corresponding to the actual environment is established. Based on the set of unhooking environment entity attribute information, the map node set is abstracted from the edge attributes. The previously obtained entity attribute information, such as the distance between each entity, passability, etc., is used to determine the connection relationship between map nodes and the attributes of the edges. If the distance between two entities is close and passable, then a connection edge can be established between the corresponding map nodes. At the same time, these connection edges are given specific attributes. For example, the length of the edge can be determined according to the actual distance between the entities, and the difficulty of the edge can be set according to the attributes of the entity (such as the size of the obstacle, crossability, etc.). Through such processing, a map connection edge set is obtained. This set describes the connection relationship and attributes between map nodes, providing a basis for subsequent map topological connections.

[0048] The map node set is topologically connected using the map connection edge set to generate a map of the dehooking environment area. Based on each node in the map node set, these nodes represent different entities in the dehooking environment. Then, based on the information in the map connection edge set, determine which nodes have a connection relationship. If there is a connection edge between two nodes, it means that there is a certain association between the corresponding entities in the actual environment, which is a passable path or a specific spatial relationship. For node pairs with connection edges, the connection operation is performed according to the attributes of the connection edge. For example, if the connection edge indicates that there is a passable road between the two entities, then the corresponding nodes are connected with corresponding lines on the map to indicate the existence of this path. The attributes of the connection edge, such as length, difficulty of passage, etc., can be reflected in different ways on the map. For example, the thickness of the line can indicate the difficulty of passage, and a longer line may correspond to a longer actual distance. By performing such a connection operation on all nodes, a complete map structure is gradually constructed. This map clearly shows the relationship between the various entities in the dehooking environment and information such as the passable path, and finally generates a map of the dehooking environment area. This map provides an important reference for subsequent path planning and the actions of the dehooking robot.

[0049] In a possible implementation, step S300 further includes:

[0050] Step S350: Mark the path obstacles in the unhooking environment area map according to the unhooking environment entity attribute information set to obtain a usable unhooking environment area map.

[0051] Step S360: define the hook removal control target, construct a path cost function according to the hook removal control target, and perform robot execution analysis according to the hook space coordinate information to determine the robot execution space coordinate information.

[0052] Step S370: Obtain a list of path planning algorithms, and use the path cost function to perform iterative planning and optimization on the robot's execution space coordinate information based on the list of path planning algorithms and the available unhooking environment area map to obtain multiple unhooking path information.

[0053] Step S380: Perform feasibility assessment and cost comparison on the multiple unhooking path information to determine the target unhooking path.

[0054] Specifically, according to the unhooking environment entity attribute information set, the path obstacles in the unhooking environment area map are marked. Since the unhooking environment entity attribute information set has been obtained in the previous step, which contains the attributes of each entity, such as passability, etc. Through these attribute information, obstacles that will hinder the progress of the unhooking robot in the map can be accurately identified, and these obstacles are marked on the unhooking environment area map, so that subsequent path planning can avoid these impassable areas. After marking, a usable unhooking environment area map is obtained, which provides a clear map environment for subsequent path planning without obstacle interference.

[0055] First, define the unhooking control target. This target is determined according to the specific application scenario and requirements. For example, it can be the shortest path to reach the hook position to save time, or minimize energy consumption to improve the robot's work efficiency. According to the determined unhooking control target, the path cost function is constructed. If the target is the shortest path, the path cost function considers the length of the path and takes the sum of the distances between the nodes on the path as the cost. If the goal is to minimize energy consumption, the path cost function is constructed by combining factors such as the robot's motion resistance on different terrains and the energy consumption required to perform different actions. At the same time, the robot execution analysis is performed based on the hook space coordinate information. This step needs to consider factors such as the robot's structural size and torque. According to the robot's structural size, the robot's movable range in different positions and postures is analyzed to ensure that the robot will not be unable to reach certain areas due to its own size limitations when performing the unhooking task. Through torque analysis, the torque required for the robot to perform actions at different positions is determined to determine whether the robot has enough power to complete the unhooking operation. The hook space coordinate information is used to determine the hook removal execution coordinates. The robot's structural dimensions and torque analysis results are comprehensively considered to select the coordinate positions suitable for the robot to perform the hook removal operation from the hook space coordinates to ensure that the robot can stably perform the hook removal task at these determined coordinates without being unable to complete the operation due to insufficient space or insufficient power. In this way, the robot's execution space coordinate information is obtained, which provides a clear target position and feasible operation space for subsequent path planning and hook removal operations.

[0056] First, obtain a list of path planning algorithms, which contains a variety of different path planning algorithms, such as the A* algorithm, the Dijkstra algorithm, and the RRT (rapidly expanding random tree) algorithm. Then, using the previously constructed path cost function, combined with the path planning algorithm list and the available hook removal environment area map, the robot performs iterative planning and optimization of the spatial coordinate information. For each planning algorithm in the list, the goal is to find the shortest and safest optimal hook removal path. Taking the A* algorithm as an example, it will calculate the cost of each node according to the path cost function, and select the node with the smallest cost for expansion. In the available hook removal environment area map, the algorithm will start from the current position of the robot and gradually search for the target node (the coordinate corresponding to the hook position), while considering factors such as the length and safety of the path, and find the optimal path from the current position to the target position through continuous iterative calculation. The Dijkstra algorithm starts from the starting node and gradually expands to other nodes, calculating the shortest distance from each node to the starting node. In this process, the obstacle information and path cost function in the available hook removal environment area map will also be used to ensure that the found path is safe and shortest. The RRT algorithm quickly explores space by randomly sampling and expanding trees. It randomly generates some points in the available unhooking environment area map, and then connects these points to the existing tree structure, gradually expanding the scope of the tree until the target location is found. Similarly, in this process, the generated path is ensured to be the shortest and safest based on the path cost function and obstacle information in the map. By optimizing different path planning algorithms, multiple unhooking path information can be obtained. These path information are the optimal paths found by each algorithm after considering factors such as path length and safety, providing multiple possibilities for subsequent path selection.

[0057] The feasibility of multiple unhooking path information is evaluated. This process is mainly to screen out the paths that the robot can actually pass. For each unhooking path, check whether it meets the robot's motion constraints. For example, considering the robot's maximum turning angle, if there is a turning operation in a certain path that requires the robot to exceed its maximum turning angle, then this path is judged to be infeasible. Similarly, check the slope of the path to ensure that the slope on the path does not exceed the robot's maximum climbing limit. You can also consider the size limit of the robot to ensure that there is enough space on the path for the robot to pass without collision. After the feasibility evaluation, a set of paths that the robot can pass is obtained, and then these passable paths are compared and optimized. The basis for the cost comparison is the path cost function constructed previously. If the path cost function mainly considers the path length, then the path with the shortest length is selected; if other factors such as time and energy consumption are also considered, these factors can be combined to calculate the total cost of each path, and then the path with the smallest total cost is selected. For example, for two traversable paths A and B, if path A is shorter but contains some complex terrain that takes more time; path B is slightly longer but has a flatter terrain, and the robot consumes less energy when passing through it, the total cost of path A and path B is calculated based on the path cost function after comprehensively considering factors such as length, time, and energy consumption, and finally the path with the smallest total cost is selected as the target hook removal path. Through such a feasibility evaluation and cost comparison optimization process, it can be ensured that the determined target hook removal path not only meets the robot's motion constraints, but also achieves the best cost, thereby providing the best action route for the robot to complete the hook removal task efficiently and accurately.

[0058] In a possible implementation, step S360 further includes:

[0059] Step S361: Acquire the structural characteristic parameters of the hook-removing robot, perform robot execution space analysis on the hook space coordinate information based on the structural characteristic parameters, and obtain the robot executable space.

[0060] Step S362: Perform torque test fitting and dynamic modeling on the unhooking robot, construct a robot dynamic model, and define an execution objective function based on the robot dynamic model.

[0061] Step S363: using the execution objective function to perform global coordinate optimization in the robot executable space to determine the robot execution space coordinate information.

[0062] Specifically, the structural characteristic parameters of the hook-removing robot are obtained. These parameters are crucial to determining the executable space of the robot, including the angle range of each joint of the robot, which determines the different postures that the robot's limbs can achieve. For example, some joints have certain rotation angle restrictions, which will affect the overall range of motion of the robot. At the same time, it also includes relevant parameters of the end position, such as the end effector, such as the structural size and length of the robot gripper. The size of the robot gripper determines the size range of objects it can grab, while the length affects the distance it can reach. Based on these structural characteristic parameters, the robot performs spatial analysis on the hook space coordinate information, considering the spatial coordinate position of the hook, combined with the structural characteristics of the robot, to determine whether the robot can reach the hook position under its own structural limitations. If a certain joint angle combination of the robot can enable the end effector to reach the spatial coordinates of the hook, then this position is within the executable space of the robot. By analyzing different joint angle combinations and considering factors such as the size and length of the end effector, the spatial range that the robot can reach can be gradually determined to meet the movable space of the robot joint to perform the hook-removing operation.

[0063] The torque test fitting of the hook-removing robot is carried out. By carrying out actual torque tests on the robot under different working conditions, a large amount of data on the torque required by the robot at different joint angles is collected. Then, the least squares method is used to process these data to establish a relationship model between torque and joint angle. Then, dynamic modeling is carried out. The robot dynamics model is constructed by comprehensively considering factors such as the robot's physical structure, mass distribution, and joint motion. This model can accurately describe the robot's motion state and torque requirements at different joint angles. For example, according to the robot's connecting rod length, mass, and joint moment of inertia, the acceleration and torque required by the robot in different postures can be calculated. Based on the constructed robot dynamics model, the execution objective function is defined. The objective function is set as the sum of joint torques, and the situation of minimizing torque is determined by optimization. When planning the robot's motion path, the robot's joint angle combination is adjusted so that the total joint torque required by the robot is minimized when reaching the hook space coordinate information, which can ensure that the robot is the most labor-saving when performing the hook-removing task.

[0064] Clarify the scope of the robot's executable space, which is determined based on the robot's structural characteristic parameters and the analysis of the hook space coordinate information. Next, define the execution objective function, such as minimizing the sum of joint torques to ensure that the robot can remove the hook with the least effort when it reaches the hook space coordinate information. Using the simulated annealing algorithm, randomly generate an initial coordinate point in the robot's executable space as the current solution, calculate the objective function value corresponding to the current solution, that is, the sum of the joint torques when the robot reaches the hook space coordinates in the current state, and then randomly generate a new coordinate point near the current solution as a candidate solution, and also calculate the objective function value of the candidate solution. If the objective function value of the candidate solution is worse than the current solution, then accept the candidate solution as the new current solution. If the objective function value of the candidate solution is not as good as the current solution, then accept the candidate solution with a certain probability. This probability gradually decreases as the algorithm proceeds, just like the cooling process in physics. In the iterative process of the algorithm, the above steps are repeated continuously, that is, new candidate solutions are generated near the current solution, evaluated and accepted. By continuously exploring and accepting a certain degree of "poor solutions", the simulated annealing algorithm can avoid falling into the local optimal solution, thereby performing global optimization in the robot's executable space. Finally, when the algorithm meets the stopping condition, the current solution is the determined robot execution space coordinate information, which enables the robot to reach the hook space coordinate in a better way to perform the hook removal operation.

[0065] In a possible implementation, step S400 further includes:

[0066] Step S410: Performing posture simulation on the unhooking robot according to the target unhooking path to obtain spatial position information of the end effector of the unhooking robot.

[0067] Step S420: Based on the dynamic characteristic parameters and motion constraint information, a movable space analysis is performed on the spatial position information of the end effector to construct a robot motion control space.

[0068] Step S430: performing hook-off execution analysis on the hook posture information based on the robot motion control space to obtain the hook-off control parameters.

[0069] Specifically, first, the target hook removal path that has been determined is clarified. This path is obtained through calculation and optimization, and is intended to provide the best action route for the hook removal robot to complete the hook removal task. Then, posture simulation is performed. Computer simulation technology or the robot's kinematic model is used to simulate the process of the robot moving along the target hook removal path. During the simulation, the robot's posture at different positions is gradually calculated based on the robot's structural characteristics, joint movement mode, and various position points on the target hook removal path. The robot's end effector is the key part that directly interacts with the hook to complete the hook removal operation. By simulating the robot's movement on the target hook removal path, the spatial position of the end effector is continuously tracked and calculated. Taking into account factors such as the robot's joint angle changes and the length of the connecting rod, the kinematic equations and geometric calculation methods are used to determine the specific coordinates of the end effector in three-dimensional space.

[0070] First, the dynamic characteristic parameters and motion constraint information of the hook-removing robot are analyzed. The dynamic characteristic parameters include the robot's motor power, torque output capacity, acceleration performance, etc. These parameters determine the power and motion capacity that the robot can generate in different states. The motion constraint information covers the robot's joint motion range limit, maximum speed limit, acceleration limit, and physical structural limits. These parameters and information are used to analyze the spatial position information of the end effector. For each spatial position of the end effector, the dynamic characteristic parameters are combined to determine whether the robot has enough power to drive itself to the position. For example, if a certain position requires the robot to output a large torque to overcome resistance, but the robot's dynamic characteristics cannot meet it, then this position is not within the range of the movable space. At the same time, considering the motion constraint information, check whether the position exceeds the robot's joint motion range or speed and acceleration limits. If these limits are exceeded, it will also be excluded from the movable space. By performing such an analysis on each spatial position of the end effector, the spatial range that the robot can reach and operate safely and effectively is gradually determined, thereby constructing the robot motion control space. This space provides a feasible motion area for subsequent unhooking operations, ensuring that the robot will not malfunction or be unable to complete the task due to exceeding its capabilities or violating motion constraints.

[0071] Based on the robot motion control space that has been constructed, this space is determined after comprehensively considering the robot's dynamic characteristic parameters and motion constraint information. It limits the range in which the robot can move safely and effectively when performing the hook removal task. Then, the hook posture information is introduced. The hook posture information includes the position, direction and specific state of the hook in space. The robot motion control space is used to perform hook removal and execution analysis on the hook posture information. In this process, it is necessary to analyze how the robot adjusts its own posture and movement in the motion control space to achieve accurate grasping of the hook. Determine the robot control parameters when grasping the hook. These parameters include the rotation angle and direction of the robot's hook removal control. By analyzing the position and direction of the hook, combined with the current position of the robot's end effector and the limitations of the motion control space, calculate the rotation angle and direction adjustment that the robot needs to make, so that the end effector can accurately align with the hook and perform the grasping operation.

[0072] In a possible implementation, step S500 further includes:

[0073] Step S510: performing hook unhooking control monitoring on the hook unhooking robot based on the target hook unhooking path and the hook unhooking control parameters to obtain hook unhooking control feedback parameters.

[0074] Step S520: performing deviation calculation on the hook-off control feedback parameter and the preset hook-off state to determine the hook-off state deviation parameter.

[0075] Step S530: Optimizing the direction of the dehooking state deviation parameter, obtaining a control parameter fine-tuning strategy, and performing deviation fine-tuning compensation on the dehooking control parameter based on the control parameter fine-tuning strategy.

[0076] Specifically, the unhooking robot is comprehensively monitored for unhooking control based on the target unhooking path and unhooking control parameters. The target unhooking path provides the robot with the best route planning to reach the hook position, while the unhooking control parameters specify various key parameters of the robot when performing the unhooking action, such as rotation angle and direction. During the monitoring process, various feedback parameters are obtained in real time through a series of sensors installed on the unhooking robot. The position sensor can accurately determine the position coordinates of the robot in space to understand whether it moves in the correct position according to the target unhooking path. The speed sensor can measure the movement speed of the robot to ensure that its speed is within the safe and efficient range. The acceleration sensor monitors the acceleration change during the robot's movement, which is very important for judging whether the robot's motion state is stable and whether the control strategy needs to be adjusted. The posture sensor can feedback the robot's body posture, including information such as tilt angle and rotation angle, to ensure that the robot maintains the correct posture when performing the unhooking task and avoid affecting the accuracy of unhooking due to improper posture. In addition, force conditions are also one of the important feedback parameters. By installing force sensors on the end effector or key parts of the robot, the force applied to the robot when it interacts with the environment and performs hook removal operations can be detected. This helps determine whether the robot collides with obstacles, whether the hook is successfully grabbed, and whether the grabbing force is appropriate. By monitoring these parameters, comprehensive hook removal control feedback parameters can be obtained, providing accurate data basis for subsequent deviation calculations and control parameter adjustments.

[0077] The two key elements of hook removal control feedback parameters and preset hook removal state are clarified. The hook removal control feedback parameters are the actual state information obtained by monitoring the hook removal robot, including position, speed, acceleration, posture, force, etc. The preset hook removal state is the state that the robot should reach according to the ideal hook removal task requirements. When calculating the deviation, each feedback parameter dimension is compared with the corresponding parameter in the preset state. For example, for the position parameter, the distance difference between the actual position and the preset position is calculated; for the speed parameter, the size difference between the actual speed and the preset speed is compared; for the posture parameter, the angle deviation between the actual posture and the preset posture is analyzed. By calculating each parameter dimension one by one, the hook removal state deviation parameters can be determined. These deviation parameters fully reflect the gap between the current actual state of the robot and the ideal preset state. If the deviation is large, it means that the actual operating state of the robot deviates greatly from the expected target and needs to be adjusted significantly; if the deviation is small, only fine-tuning is needed to make the robot reach the preset hook removal state. Determining the decoupling state deviation parameters provides a specific quantitative basis for subsequent optimization direction analysis and control parameter fine-tuning, so that the robot's control parameters can be adjusted in a targeted manner to reduce the deviation and enable the robot to perform the decoupling task more accurately.

[0078] The optimization direction of the dehooking state deviation parameters is analyzed. Since the deviation parameters reflect the gap between the actual state of the robot and the preset dehooking state, by analyzing these deviations, the direction that needs to be adjusted can be determined to reduce the deviation. The analysis methods are different for different types of deviation parameters. For example, if the position deviation shows that the actual position of the robot deviates from the preset position, it is necessary to determine in which direction the robot should move to reduce the position deviation. If the speed deviation indicates that the robot is too fast or too slow, it is necessary to determine whether to increase or decrease the speed to meet the preset speed requirements. For posture deviation, analyze which joint angles need to be adjusted to make the robot's posture closer to the preset posture. Through such analysis, the control parameter fine-tuning strategy is obtained, which specifically points out the direction and amplitude of adjusting the dehooking control parameters. For example, if the position deviation is large, the strategy is to adjust the robot's movement speed and direction to make it move faster toward the preset position; if the posture deviation is small, the strategy is to fine-tune the angles of certain joints to gradually correct the robot's posture. Based on the control parameter fine-tuning strategy, the dehooking control parameters are fine-tuned for deviation compensation, and the dehooking control parameters are fine-tuned according to the adjustment direction and amplitude determined in the strategy. For example, if the strategy requires increasing the rotation angle of a certain joint of the robot, the angle value of the joint in the hook removal control parameter will be increased accordingly. By continuously performing deviation fine-tuning compensation, the hook removal state deviation can be gradually reduced, making the actual state of the robot closer and closer to the preset hook removal state, thereby improving the accuracy and stability of the hook removal operation.

[0079] Embodiment 2 is based on the same inventive concept as the dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism in the above embodiment. Figure 2 As shown, the present application provides a dynamic hook removal control system for a hook removal robot based on a feedback mechanism, and the embodiments of the present application and the method embodiments are based on the same inventive concept. The system includes:

[0080] The image information acquisition module 10 is used to acquire a hook-removing robot, which integrates a visual sensor and an image processing unit. The hook-removing robot is moved to a preset hook-removing position, and the visual sensor is activated at the same time to collect and obtain hook-removing image information and environmental perception image information.

[0081] The hook spatial coordinate information determination module 20 performs edge detection and feature extraction on the hook removal image information based on the image processing unit to obtain the hook position information and hook posture information, and performs coordinate conversion on the hook position information according to the preset hook removal position to determine the hook spatial coordinate information.

[0082] The target unhooking path acquisition module 30 is used to perform feature abstraction on the environmental perception image information, generate an unhooking environment area map, perform path planning on the hook space coordinate information based on the unhooking environment area map, and obtain the target unhooking path.

[0083] The hook removal control parameter acquisition module 40 is used to obtain the power characteristic parameters and motion constraint information of the hook removal robot, and based on the power characteristic parameters and motion constraint information, the hook posture information is analyzed to obtain the hook removal control parameters.

[0084] A dynamic unhooking control module 50 performs dynamic unhooking control on the unhooking robot based on the target unhooking path and the unhooking control parameters, and introduces a feedback mechanism to perform real-time deviation analysis and compensatory unhooking control on the unhooking control process.

[0085] Furthermore, the hook space coordinate information determination module 20 also includes:

[0086] A hook-removing standard grayscale image acquisition unit is used to convert the hook-removing image information into a hook-removing grayscale image, and filter, denoise and enhance the hook-removing grayscale image to obtain a hook-removing standard grayscale image.

[0087] A hook removal edge detection information acquisition unit is used to perform edge detection on the hook removal standard grayscale image through a Canny edge detector to obtain hook removal edge detection information.

[0088] The hook removal contour information acquisition unit is used to preset a low edge threshold and a high edge threshold, and based on the low edge threshold and the high edge threshold, perform edge contour screening on the hook removal edge detection information to obtain the hook removal contour information.

[0089] A hook position information acquisition unit is used to perform anchor frame positioning on the hook removal contour information to obtain hook removal anchor frame information, perform center point calculation based on the hook removal anchor frame information, and obtain the hook position information.

[0090] A hook posture information acquisition unit is used to extract hook posture features based on the hook removal contour information and the hook position information to obtain the hook posture information.

[0091] Furthermore, the target unhooking path acquisition module 30 also includes:

[0092] A multi-dimensional feature set acquisition unit for a hook removal environment is used to perform multi-dimensional feature extraction and feature dimensionality reduction processing on the environmental perception image information to obtain a multi-dimensional feature set of a hook removal environment.

[0093] A unit for acquiring an entity attribute information set of an unhooking environment. The unit for acquiring an entity attribute information set of an unhooking environment performs distributed entity recognition based on the multi-dimensional feature set of the unhooking environment to obtain an entity set of the unhooking environment, and performs attribute marking on the entity set of the unhooking environment to obtain an entity attribute information set of the unhooking environment.

[0094] A map connection edge set acquisition unit is used to abstract the unhooking environment entity set into a map node set, perform edge attribute abstraction on the map node set based on the unhooking environment entity attribute information set, and obtain a map connection edge set.

[0095] A hook-off environment area map generation unit is provided, wherein the hook-off environment area map generation unit performs a map topology connection on the map node set based on the map connection edge set to generate the hook-off environment area map.

[0096] Furthermore, the target unhooking path acquisition module 30 also includes:

[0097] An available unhooking environment area map acquisition unit is used to mark the path obstacles in the unhooking environment area map according to the unhooking environment entity attribute information set to obtain the available unhooking environment area map.

[0098] An execution space coordinate information determination unit is used to define a hook removal control target, construct a path cost function according to the hook removal control target, and perform robot execution analysis according to the hook space coordinate information to determine the robot execution space coordinate information.

[0099] A plurality of hook-removing path information acquisition units are used to obtain a list of path planning algorithms, and to perform iterative planning and optimization on the robot's execution space coordinate information based on the list of path planning algorithms and the available hook-removing environment area map using the path cost function to obtain a plurality of hook-removing path information.

[0100] A target unhooking path determination unit is used to perform feasibility assessment and cost comparison and optimization on the multiple unhooking path information to determine the target unhooking path.

[0101] Furthermore, the execution space coordinate information determination unit further includes:

[0102] A robot executable space acquisition unit is used to acquire the structural characteristic parameters of the unhooking robot, and perform robot execution space analysis on the hook space coordinate information based on the structural characteristic parameters to obtain the robot executable space.

[0103] An execution objective function definition unit is used to perform torque test fitting and dynamic modeling on the unhooking robot, build a robot dynamic model, and define an execution objective function based on the robot dynamic model.

[0104] A coordinate global optimization unit is used to use the execution objective function to perform coordinate global optimization in the robot executable space to determine the robot execution space coordinate information.

[0105] Furthermore, the unhooking control parameter acquisition module 40 further includes:

[0106] A spatial position information acquisition unit is used to perform posture simulation on the hook removal robot according to the target hook removal path to obtain the spatial position information of the end effector of the hook removal robot.

[0107] A motion control space construction unit is used to perform a movable space analysis on the spatial position information of the end effector based on the dynamic characteristic parameters and the motion constraint information to construct a robot motion control space.

[0108] A hook unhooking execution analysis unit is used to perform hook unhooking execution analysis on the hook posture information based on the robot motion control space to obtain the hook unhooking control parameters.

[0109] Furthermore, the dynamic unhooking control module 50 further includes:

[0110] A hook unhooking control feedback parameter acquisition unit is used to monitor the hook unhooking control of the hook unhooking robot based on the target hook unhooking path and the hook unhooking control parameters to obtain the hook unhooking control feedback parameters.

[0111] A hook-unhooking state deviation parameter determination unit is used to perform deviation calculation on the hook-unhooking control feedback parameter and the preset hook-unhooking state to determine the hook-unhooking state deviation parameter.

[0112] A deviation fine-tuning compensation unit is used to perform optimization direction analysis on the decoupling state deviation parameter, obtain a control parameter fine-tuning strategy, and perform deviation fine-tuning compensation on the decoupling control parameter based on the control parameter fine-tuning strategy.

[0113] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0114] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0115] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A dynamic hook removal control method for a hook removal robot based on a feedback mechanism, characterized in that: The method comprises: Acquire a hook-removing robot, wherein the hook-removing robot integrates a visual sensor and an image processing unit, move the hook-removing robot to a preset hook-removing position, and simultaneously activate the visual sensor to collect and obtain hook-removing image information and environment perception image information; Based on the image processing unit, edge detection and feature extraction are performed on the hook removal image information to obtain hook position information and hook posture information, and coordinate conversion is performed on the hook position information according to the preset hook removal position to determine the hook space coordinate information; Performing feature abstraction on the environmental perception image information to generate a hook removal environment area map, and performing path planning on the hook space coordinate information based on the hook removal environment area map to obtain a target hook removal path; Acquiring the power characteristic parameters and motion constraint information of the hook-removing robot, and performing hook-removing analysis on the hook posture information based on the power characteristic parameters and motion constraint information to obtain hook-removing control parameters; The unhooking robot is dynamically controlled based on the target unhooking path and the unhooking control parameters, and a feedback mechanism is introduced to perform real-time deviation analysis and compensatory unhooking control on the unhooking control process.

2. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 1 is characterized in that: The obtaining of the hook position information includes: Converting the hook removal image information into a hook removal grayscale image, and performing filtering, denoising and image enhancement on the hook removal grayscale image to obtain a hook removal standard grayscale image; Performing edge detection on the hook removal standard grayscale image by using a Canny edge detector to obtain hook removal edge detection information; Preset an edge low threshold and an edge high threshold, and perform edge contour screening on the hook removal edge detection information based on the edge low threshold and the edge high threshold to obtain hook removal contour information; Anchor frame positioning is performed on the hook removal contour information to obtain hook removal anchor frame information, and center point calculation is performed based on the hook removal anchor frame information to obtain the hook position information; The hook posture feature is extracted based on the hook removal contour information and the hook position information to obtain the hook posture information.

3. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 1, characterized in that: The generating of the decoupling environment area map comprises: Performing multi-dimensional feature extraction and feature dimension reduction processing on the environmental perception image information to obtain a multi-dimensional feature set of the hook removal environment; Perform distributed entity recognition based on the multi-dimensional feature set of the unhooking environment to obtain an unhooking environment entity set, and perform attribute marking on the unhooking environment entity set to obtain an unhooking environment entity attribute information set; Abstracting the unhooking environment entity set into a map node set, performing edge attribute abstraction on the map node set based on the unhooking environment entity attribute information set, and obtaining a map connection edge set; The map node set is topologically connected based on the map connection edge set to generate the dehooking environment area map.

4. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 3 is characterized in that: The step of obtaining the target hook removal path comprises: Marking the path obstacles in the dehooking environment area map according to the dehooking environment entity attribute information set to obtain an available dehooking environment area map; Defining a hook removal control target, constructing a path cost function according to the hook removal control target, and performing robot execution analysis according to the hook space coordinate information to determine the robot execution space coordinate information; Obtain a path planning algorithm list, and use the path cost function to perform iterative planning and optimization on the robot execution space coordinate information based on the path planning algorithm list and the available hook removal environment area map to obtain multiple hook removal path information; Perform feasibility assessment and cost comparison on the multiple unhooking path information to determine the target unhooking path.

5. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 4 is characterized in that: The determining of the robot execution space coordinate information includes: Acquire the structural characteristic parameters of the hook-removing robot, perform robot execution space analysis on the hook space coordinate information based on the structural characteristic parameters, and obtain the robot executable space; Performing torque test fitting and dynamic modeling on the unhooking robot, building a robot dynamic model, and defining an execution objective function according to the robot dynamic model; The execution objective function is used to perform global coordinate optimization in the robot executable space to determine the robot execution space coordinate information.

6. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 1, characterized in that: The obtaining of the unhooking control parameters includes: Performing posture simulation on the hook-removing robot according to the target hook-removing path to obtain spatial position information of the end effector of the hook-removing robot; Based on the dynamic characteristic parameters and motion constraint information, a movable space analysis is performed on the spatial position information of the end effector to construct a robot motion control space; The hooking posture information is analyzed based on the robot motion control space to obtain the hooking control parameters.

7. The dynamic hook-removing control method of the hook-removing robot based on the feedback mechanism as claimed in claim 1, characterized in that: The feedback mechanism is introduced to perform real-time analysis of deviations and compensation for hook removal control during the hook removal control process, including: Based on the target hook removal path and the hook removal control parameters, the hook removal robot is monitored for hook removal control to obtain a hook removal control feedback parameter; Performing deviation calculation on the hook removal control feedback parameter and the preset hook removal state to determine the hook removal state deviation parameter; The decoupling state deviation parameter is analyzed in an optimization direction to obtain a control parameter fine-tuning strategy, and the decoupling control parameter is fine-tuned and compensated for the deviation based on the control parameter fine-tuning strategy.

8. The dynamic hook removal control system of the hook removal robot based on feedback mechanism is characterized by: The system comprises: An image information acquisition module, wherein the image information acquisition module is used to acquire a hook-removing robot, wherein the hook-removing robot integrates a visual sensor and an image processing unit, moves the hook-removing robot to a preset hook-removing position, and simultaneously activates the visual sensor to acquire hook-removing image information and environment perception image information; A hook space coordinate information determination module, wherein the hook space coordinate information determination module performs edge detection and feature extraction on the hook removal image information based on the image processing unit to obtain hook position information and hook posture information, and performs coordinate conversion on the hook position information according to the preset hook removal position to determine the hook space coordinate information; A target hook removal path acquisition module, the target hook removal path acquisition module is used to perform feature abstraction on the environmental perception image information, generate a hook removal environment area map, perform path planning on the hook space coordinate information based on the hook removal environment area map, and obtain a target hook removal path; A hook removal control parameter acquisition module, the hook removal control parameter acquisition module is used to obtain the power characteristic parameters and motion constraint information of the hook removal robot, and based on the power characteristic parameters and motion constraint information, perform hook removal analysis on the hook posture information to obtain the hook removal control parameters; A dynamic unhooking control module is provided, wherein the dynamic unhooking control module performs dynamic unhooking control on the unhooking robot based on the target unhooking path and the unhooking control parameters, and introduces a feedback mechanism to perform real-time deviation analysis and compensatory unhooking control on the unhooking control process.

Images