Robot active decoupling aerial cable grabbing method based on topological segmentation

Abstract

The application discloses a robot active decoupling type aviation cable grabbing method based on topological segmentation, and specifically comprises the following steps: a vision camera acquires position information of each cable connector in a cable bundle; an aviation cable cross-section geometric analysis model is constructed; a topological segmentation path set of the aviation cable bundle is planned; path pruning is performed according to multi-level constraints such as target avoidance, grabbing quantity, safety gap target exposure angle and the like; optimal segmentation paths are obtained through multi-objective optimization decision making by taking the grabbing quantity, safety crossing difficulty and shielding influence as optimization indexes; and a double-arm robot actively decouples to grab a target cable. The method provided by the application can realize reliable grabbing of a target to be operated under dense shielding of aviation cables, provides an effective scheme for dexterous operation of cables in a robot cable assembly process, and thus promotes the development of cable automatic assembly technology in the field of aviation and aerospace.

Description

Technical Field

[0001] This invention relates to the field of flexible cable grasping operations for aerospace robots, and more particularly to a robot-based active decoupling method for grasping aerospace cables based on topology segmentation. Background Technology

[0002] Aviation cables play a crucial role in the transmission of power, signal control, and data exchange within aircraft. Their layout within the fuselage is becoming increasingly dense and complex, and their assembly quality directly impacts flight safety and overall aircraft performance. Traditional manual assembly methods suffer from bottlenecks such as high labor intensity, low efficiency, and poor quality consistency, making it difficult to meet the urgent demands of modern aerospace manufacturing for high precision, high quality, and high production capacity. Automated assembly of aviation cables using robots can significantly reduce the workload of workers, greatly improve assembly efficiency and quality consistency, and is of great significance for promoting the upgrading of the aerospace manufacturing industry.

[0003] In actual assembly sites, aviation cables are often suspended in bundles. Unlike rigid objects, aviation cables have complex flexible deformation characteristics. Under the influence of gravity, they are densely clustered when naturally suspended, making it easy for the target cable to be obstructed by the outer cables, thus preventing direct grasping. Existing target grasping methods under obstruction conditions mainly focus on rigid objects. The common obstacle avoidance strategy of removing obstacles one by one faces serious challenges in handling such problems: due to the lack of an intermediate holding mechanism for the displaced cable in the environment, once the robot releases a single obstacle cable, it will quickly return to its initial equilibrium position under the coupling effect of gravity and its own bending stiffness, causing the target cable to be "re-obstructed". Therefore, there is an urgent need for a target grasping method that can handle flexible cables with obstacle obstruction, in order to improve the robot's ability to manipulate flexible cables. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a robot-based active decoupled aviation cable grasping method based on topology segmentation. By comprehensively considering the influencing factors of obstacles and performing multi-objective optimization, an optimal topology segmentation path is planned. An active decoupling method is used to remove the obstacle cable clusters that obstruct the target cable in one go, aiming to successfully expose the target cable for accurate grasping, thereby significantly improving the robot's ability to operate in complex and flexible environments.

[0005] To achieve the above-mentioned technical objectives, the present invention provides the following technical solution: A robot-based active decoupled aviation cable grasping method based on topology segmentation specifically includes: S1. Obtain the position and radius information of each aviation cable in the suspended aviation cable bundle; the aviation cable is composed of flexible cable and cable connector; S2. Set the capture safety threshold, combine the location information and radius information of the aviation cable to construct the cross-sectional geometric analysis model of the aviation cable, and mark the target cable and add obstacle constraints; S3. Based on the position and geometric relationship between each aviation cable in the cross-sectional geometric analysis model, construct a topology segmentation path set; S4. Based on the vector field histogram algorithm, quantify the independent occlusion intensity of each obstacle cable on the target cable; S5. Prune the topology segmentation path according to multi-level constraints including target avoidance constraints, capture quantity constraints, safety gap constraints, and target exposure angle constraints, and retain the topology segmentation path that satisfies the multi-level constraints; execute step S6; if there is no topology segmentation path that satisfies all four constraints after pruning, return to step S1 and start execution again after randomly moving the aviation cable bundle. S6. Considering optimization indicators including the number of captures, the difficulty of safe traversal, and the impact of occlusion, perform multi-objective optimization decision-making on the topology segmentation path to obtain the optimal segmentation path; S7. The dual-arm robot uses an active decoupling method to collaboratively complete the grasping operation based on the optimal segmentation path.

[0006] Furthermore, step S2 specifically involves: Calculate the horizontal cross-sectional height suitable for the robot's active decoupling operation, the horizontal cross-sectional height It is obtained by adding the maximum height of the cable connector center, half of the maximum length of the cable connector, the maximum length of the cable connector tail accessory, and the preset cross-sectional safety margin; Constructing a cross-sectional geometric analysis model based on top-view projection: taking Using the plane as the analysis section, the spatial distribution of each aviation cable is projected onto the analysis section. Based on the horizontal and vertical coordinates of the center of each aviation cable, concentric circles with the radius of the cable connector, the radius of the flexible cable, and the gripping safety threshold are drawn on the analysis section, respectively, and are denoted as the cable connector circle, the flexible cable circle, and the gripping safety threshold circle. These concentric circles are used as simplified projections of each aviation cable on the analysis section. Record the serial number of each aviation cable and mark the target cable; at the same time, based on the positional relationship between external obstacles and aviation cable bundles, simplify external obstacles into straight line boundaries on the cross section, as obstacle constraints for subsequent topology analysis.

[0007] Furthermore, the gripping safety threshold mentioned in step S2 is composed of the radius of the aviation flexible cable, the product of the cable radial manipulation factor and the radius of the aviation flexible cable, the half width of a single clamping piece of the gripper, the system error margin, and the preset safety margin; the cable radial manipulation factor is determined based on the flexible deformation characteristics of the aviation cable.

[0008] Furthermore, step S3 specifically includes: S31. In the cross-sectional geometric analysis model, select any two aviation cables. Using the grab safety threshold circles of these two aviation cables as a reference, construct a combined ellipse whose major axis is collinear with the line connecting the centers of the two grab safety threshold circles and whose foci are the centers of the two grab safety threshold circles. The combined ellipse satisfies the constraint of including the grab safety threshold circles and the boundary safety margin. The semi-major axis of the combined ellipse is determined by the distance between the centers of the two grab safety threshold circles and the larger radius of the two grab safety threshold circles. The semi-minor axis of the combined ellipse is determined by the semi-major axis and the preset shape adjustment ratio constant. The constraint of including the grab safety threshold circles and the boundary safety margin is that the combined ellipse must completely include the two grab safety threshold circles, and the shortest distance from any point on the combined ellipse to the two grab safety threshold circles is not less than the preset safety margin. S32. Draw perpendicular lines to the major axis of the combined ellipse through the centers of the flexible cables of the two aviation cables respectively. Take the intersection of the perpendicular line and the combined ellipse as the tangent point. Draw tangent lines to the combined ellipse based on the tangent point. The continuous path that connects the tangent segment, the ellipse arc segment, and the tangent segment in sequence is the topology segmentation path. The topology segmentation path divides the analysis section into a grab cluster located inside the topology segmentation path and a retain cluster located outside the topology segmentation path. Each combined ellipse generates two topology segmentation paths. S33. Repeat steps S31-S32 for all aviation cable combinations to generate a topology partitioning path set containing all potential partitioning schemes.

[0009] Furthermore, step S4 specifically includes: A polar coordinate system is constructed with the center of the flexible cable circle of the target cable as the origin, and the analysis section is divided into N sectors, each sector containing K virtual ray network units; the obstacle strength of each virtual ray network unit is calculated based on preset deterministic values, preset constant parameters, and the nearest Euclidean distance from the ray origin to the obstacle. The obstacle strengths of all virtual ray network units in each sector are summed to obtain the obstacle histogram value for each sector; For an obstruction cable, calculate the sum of the obstruction histogram values ​​of all sectors it occupies, which is taken as the independent blocking strength of the obstruction cable to the target cable.

[0010] Furthermore, the target avoidance constraints, capture quantity constraints, safety gap constraints, and target exposure angle constraints in step S5 are specifically as follows: Target avoidance constraint: The target cable must be segmented into the reserved cluster rather than the grab cluster; Grab quantity constraint: The number of aviation cables in the grab cluster must not exceed the upper limit of the gripper capacity; Safety clearance constraint: The shortest distance from the center of all aviation cables to the topology split path must not be less than the minimum safety distance, which is determined based on the flexible cable radius and the cable radial manipulation factor; Target exposure angle constraint: The angle of the continuous exposed sector obtained by the target cable shall not be less than the preset angle threshold; the angle of the continuous exposed sector is the maximum value of the sum of the central angles of the continuous sectors with the obstacle histogram value always being zero in the polar coordinate system after removing all obstacle cables in the grab cluster.

[0011] Furthermore, in step S6, the occlusion effect consists of the cumulative occlusion intensity of the obstacle cables in the capture cluster and the continuous exposure sector angle of the target cable.

[0012] Furthermore, in step S6, the safety crossing difficulty is composed of the number of cables traversed along the path and the spatial spacing factor, specifically: First, calculate the shortest distance from the center of any aviation cable connector to the topology split path. If satisfied If so, the cable is determined to be a path crossing cable; in, The connector radius of the aviation cable. Find the minimum radius among all cable connector radii; iterate through all aviation cables to obtain the number of cables the path crosses; Spacing factor Calculated using the following formula: ; in, , These represent the minimum and average distances between the capture safety threshold circle for all paths traversing cables and the topology segmentation path, respectively. and These are the weights of the corresponding values, and their sum is 1.

[0013] Furthermore, step S6 specifically includes: The number of samples captured, the impact of occlusion, and the difficulty of safe passage were normalized. After normalization, a global objective function is established based on a linear weighted average of three factors: the number of captures, the impact of occlusion, and the difficulty of safe traversal. The topology segmentation path with the largest global objective function is selected as the optimal segmentation path for final execution.

[0014] Furthermore, step S7 specifically includes: A dual-arm robot consisting of a gripping arm and a decoupling arm is used to perform the gripping task. First, the decoupling arm moves along the optimal segmentation path to gather, clamp, and remove the non-target aviation cable. The removal displacement meets the cable's flexible range of motion constraint. The cable's flexible range of motion constraint is that the removal displacement of the obstacle cable must not exceed the preset movable range of the aviation cable. The gripper uses a vision system to acquire the pose of the exposed target cable, plans a trajectory to move to the target point and close the gripper to complete precise gripping. According to the requirements of subsequent tasks, perform cable collaborative handover; the gripping arm and the decoupling arm cooperate in the collaborative area to transfer the target cable from the former to the end gripper of the latter.

[0015] By employing the above technical solution, the present invention provides a robot-based active decoupled aviation cable grasping method based on topology segmentation, which has at least the following beneficial effects: The method proposed in this invention effectively solves the problem of difficulty in grasping target cables in scenarios where flexible cables are obstructed. By constructing a topology segmentation path set, the complex cable bundle is divided into grasping clusters and retention clusters. A dual-arm robot is used in collaboration, with one arm actively decoupling to grasp and remove the obstructing cable clusters, creating an unobstructed target grasping space for the other arm to directly grasp the target cable. A Vector Field Histogram (VFH) algorithm is introduced to quantify the cable occlusion, transforming the vague concept of "occlusion" into calculable "obstacle strength" and "continuous gap angle." A multi-objective optimization method is established, encompassing the number of grasps, the difficulty of safe passage, and the impact of occlusion. This method plans an optimal path that maximizes the exposure of the target cable (maximum continuous gap angle) while minimizing the robot's operational difficulty (fewer grasps, lower crossing risk), significantly improving the efficiency and success rate of target cable grasping. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is an overall flowchart of the robot-based active decoupling aviation cable grasping method based on topology segmentation proposed in this invention; Figure 2 This is a schematic diagram of a dual-arm collaborative grasping operation scenario in an embodiment of the present invention; Figure 3 This is a schematic diagram of the cross-sectional geometric analysis model and topological segmentation path design in an embodiment of the present invention; Figure 4 This is a schematic diagram of the Vector Field Histogram (VFH) algorithm in an embodiment of the present invention; Figure 5This is a flowchart illustrating the selection of the optimal topology segmentation path in an embodiment of the present invention. Figure 6 This is a schematic diagram of the optimal topology partitioning path and its VFH in an embodiment of the present invention; Figure 7 This is a flowchart of a dual-robot collaborative target grasping process in an embodiment of the present invention.

[0017] Reference numerals: 1-Vision camera; 2-Left-arm collaborative robot; 3-Left-arm collaborative robot end effector; 4-Left-arm collaborative robot vision camera; 5-Aeronautical cable bundle; 6-External obstacle; 7-Right-arm collaborative robot vision camera; 8-Right-arm collaborative robot end effector; 9-Right-arm collaborative robot. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] Although the steps in this invention are arranged by reference numerals, this is not intended to limit the order of the steps. Unless the order of the steps is explicitly stated or the execution of a step requires other steps as a basis, the relative order of the steps can be adjusted. It is understood that the term "and / or" as used herein refers to and covers any and all possible combinations of one or more of the associated listed items.

[0020] This invention proposes a robot-based active decoupled aviation cable grasping method based on topology segmentation, which is used to successfully grasp target aviation cables in scenarios with obstructions. Figure 1 As shown, it specifically includes the following steps: S1. Obtain the position and radius information of each aviation cable in the suspended aviation cable bundle; the aviation cable is composed of flexible cable and cable connector.

[0021] In this embodiment, one end of the flexible cable is connected to one end of the cable connector, with their centers coinciding at the connection point. Since the flexible cable also possesses a certain degree of rigidity, a small section of the flexible cable and the cable connector at the connection point can be considered as two coaxial cylinders. When acquiring positional information visually, obtaining the pose of the cable connector is easier and more accurate than obtaining the position of the flexible cable. Therefore, by obtaining the Cartesian position of the center of each cable connector in the world coordinate system, the position of the center of the flexible cable can be obtained, which is the position information of the entire aviation cable. The radius information, namely the radius of the flexible cable and the radius of the cable connector, is determined by the different specifications of each aviation cable when it was initially purchased or manufactured, and can also be directly measured before operation.

[0022] In this embodiment, the following method was used: Figure 2 The dual-arm collaborative system shown acquires position information and performs grasping; the system is as follows: Figure 2 As shown, the system includes a vision camera 1, a left-arm collaborative robot 2, a left-arm collaborative robot end effector 3, a left-arm collaborative robot vision camera 4, an aviation cable bundle 5, an external obstacle 6, a right-arm collaborative robot vision camera 7, a right-arm collaborative robot end effector 8, and a right-arm collaborative robot 9. Vision camera 1 is placed directly below the aviation cable bundle 5, observing from below to obtain the cross-sectional position information of the cable connectors. The left-arm collaborative robot vision camera 4 and the right-arm collaborative robot vision camera 7 are installed at the ends of the two collaborative robots to obtain the pose of the aviation cables and their corresponding serial numbers. In step S1, the two robots are used to move their vision cameras to observe the aviation cable bundle around it to obtain the height and serial number of each aviation cable, and to match this information with that of vision camera 1, ultimately obtaining the position information of each aviation cable. .

[0023] S2. Set the capture safety threshold, combine the location information and radius information of the aviation cable to construct the cross-sectional geometric analysis model of the aviation cable, and mark the target cable and add obstacle constraints; In a preferred embodiment, step S2 specifically comprises: Calculate the horizontal cross-sectional height suitable for the robot's active decoupling operation, the horizontal cross-sectional height It is obtained by adding the maximum height of the cable connector center, half the maximum length of the cable connector, the maximum length of the cable connector tail accessory, and the preset cross-sectional safety margin, expressed by the formula: ; in, The largest z-axis coordinate value of all connector centers obtained from visual measurement in the world coordinate system; and These are the maximum lengths of all aviation cable connectors and tail accessories; To ensure a safety margin for the cross-section, the cross-section is for the cable section of aviation cables.

[0024] Constructing a cross-sectional geometric analysis model based on top-view projection: taking The plane is used as the analysis section (the analysis section projects all aviation cables and external obstacles). The spatial distribution of each aviation cable is projected onto the analysis section, and the horizontal and vertical coordinates of the center of each aviation cable are used as the basis for the analysis. Concentric circles with radii equal to the cable connector radius, the flexible cable radius, and the gripping safety threshold radius are drawn, denoted as the cable connector circle, the flexible cable circle, and the gripping safety threshold circle, respectively. These concentric circles are used as simplified projections of each aviation cable onto the analysis section for subsequent analysis and planning of collision-free topology segmentation paths. Figure 3 In the diagram shown, each set of concentric circles corresponds to an aviation cable. The large green circle represents the cable connector circle, the medium orange circle represents the grab safety threshold circle, and the small blue / red circle represents the cable.

[0025] It should be noted that, near the connection between the cable and the connector, the connector and the cable are approximately two coaxial cylinders, and the analysis section is generally located near the connection. Even if there is some deviation, the method proposed in this application uses a safety margin to eliminate the deviation in the subsequent setting of the capture safety threshold. Therefore, when constructing the cross-sectional geometric analysis model, it is only necessary to know the center position of each connector (i.e., the center position of the flexible cable), the connector radius, the flexible cable radius, and the capture safety threshold radius to draw concentric circles, which can then be used for subsequent topology segmentation path planning.

[0026] Furthermore, in this embodiment, the gripping safety threshold is a preset safety planning range designed to offset system errors and ensure that the topology segmentation path can effectively separate the cable bundle. It is composed of the radius of the aerospace flexible cable, the product of the cable radial manipulation factor and the radius of the aerospace flexible cable, the half-width of a single clamping piece, the system error margin, and the preset safety margin. The cable radial manipulation factor is determined based on the flexible deformation characteristics of the aerospace cable, and is expressed by the following formula: ; in, Let be the radius of the flexible cable of the i-th aviation cable; The cable radial manipulation factor defines the maximum effective contact depth ratio that the gripper is allowed to cut into the cable geometry radius when performing a splitting operation. The value is determined based on the flexible deformation characteristics and surface friction coefficient of the cable, and is usually set to 0.3~0.5; Half the width of a single gripper piece; This is the error margin; This is the set safety margin.

[0027] Record the serial number of each aviation cable and mark the target cable; simultaneously, based on the positional relationship between external obstacles and the aviation cable bundle, simplify the external obstacles into straight line boundaries on the cross-section as obstacle constraints for subsequent topology analysis; in actual operation scenarios, external obstacles may appear behind and on both sides of the suspended aviation cable bundle; this application directly simplifies the projection of the external obstacle into a straight line mainly for marking obstacles, so that when calculating the angle of continuous exposure sectors in the subsequent calculation, the ray hitting the simplified mark of the obstacle will know which area has an obstacle, and thus no exposure angle will be generated in that area, so the robot will not operate in the obstacle area. Simplifying according to the outline of the external obstacle is also possible, but simplifying to a straight line is simpler and does not affect the obstacle marking effect, such as Figure 3The target cable is shown in red, and the green line above it represents a simplified external obstacle.

[0028] S3. Based on the position and geometric relationship between each aviation cable in the cross-sectional geometric analysis model, construct a topology segmentation path set; In a preferred embodiment, step S3 specifically includes: S31. In the cross-sectional geometric analysis model, select any two aviation cables (in this embodiment, select the two aviation cables with serial numbers i and j). Using the grab safety threshold circles of these two aviation cables as the reference, construct a combined ellipse whose major axis is collinear with the line connecting the centers of the two grab safety threshold circles and whose foci are the centers of the two grab safety threshold circles. The combined ellipse satisfies the constraint of including the grab safety threshold circles and the boundary safety margin. The semi-major axis of the combined ellipse is determined by the distance between the centers of the two grab safety threshold circles and the larger radius of the two grab safety threshold circles. The semi-minor axis of the combined ellipse is determined by the semi-major axis and the preset shape adjustment ratio constant. The constraint of including the grab safety threshold circles and the boundary safety margin is that the combined ellipse must completely include the two grab safety threshold circles, and the shortest distance from any point on the combined ellipse to the two grab safety threshold circles is not less than the preset safety margin.

[0029] In this embodiment, the equation of the combined ellipse is expressed as: ; in, , D is the distance between the two centers; a is the semi-major axis of the combined ellipse. b is the semi-minor axis of the composite ellipse. m and n are shape adjustment proportional constants, and ; The larger of the two grab safety threshold radii is given. The preset safety margin is n× ; like Figure 3 As shown in the figure, the black dashed ellipse is the constructed ellipse. This combined ellipse is constructed based on aviation cables No. 7 and No. 8, and completely contains aviation cables No. 7 and No. 8 inside the combined ellipse.

[0030] S32. Draw perpendicular lines from the centers of the flexible cables of the two aviation cables to the major axis of the combined ellipse. Using the intersection of these perpendicular lines and the combined ellipse as the tangent point, draw tangent lines to the combined ellipse based on this tangent point. The continuous path connecting the tangent segments, elliptical arc segments, and tangent segments in sequence is used as the topology segmentation path. This topology segmentation path divides the analysis section into a grasping cluster (i.e., the set of obstacle cables to be grasped and removed) located inside the topology segmentation path and a retention cluster located outside the topology segmentation path. Each combined ellipse generates two topology segmentation paths; similarly... Figure 3 As shown in the diagram, the black dashed line represents the tangent boundary, and the thick red dashed line represents the topology partitioning path on the right, constructed based on cables 7 and 8 (left and right are relative to the major axis of the composite ellipse). This path designates cables 7, 8, and 9 within the shaded area as the grab cluster, and the remaining cables as the retain cluster; similarly, if the topology partitioning path on the left is selected, all cables (1-10) in this instance will be assigned to the grab cluster. S33. Repeat steps S31-S32 for all aviation cable combinations to generate a topology partitioning path set containing all potential partitioning schemes.

[0031] S4. Based on the Vector Field Histogram (VFH) algorithm, quantify the independent occlusion intensity of each obstacle cable (i.e., non-target cable) on the target cable. In a preferred embodiment, step S4 specifically includes: like Figure 4 As shown, a polar coordinate system is constructed with the center of the flexible cable circle of the target cable (red circle in the figure) as the origin (with y=60 as the obstacle boundary, shown as a red line in the figure), and the analysis section is divided into N sectors, each sector containing K virtual ray network elements. The obstacle strength of each virtual ray mesh element is calculated based on preset deterministic values, preset constant parameters, and the nearest Euclidean distance from the ray origin to the obstacle (including the obstacle cable and external obstacles). The formula is expressed as: ; in, This represents the obstacle strength of the k-th virtual ray mesh cell. This is the deterministic value of the k-th virtual ray mesh element. In this invention, the model has high determinism, so it is generally set to 1. , For constant parameters; The nearest Euclidean distance from the ray origin of the k-th virtual ray mesh cell to the obstacle (the green value shown in the figure is the average Euclidean distance from the ray origin of each sector to the nearest obstacle). The obstacle strengths of all virtual ray network units in each sector are summed to obtain the obstacle histogram value for each sector. And draw the obstacle histogram of the target cable in the current scene, such as... Figure 4 As shown, the histogram value is 0 in the range of 160° to 215° (i.e., unobstructed gaps).

[0032] For an obstructed cable, the sum of the obstruction histogram values ​​of all sectors it occupies is calculated as the independent blocking strength of the obstructed cable to the target cable. , Let q be the set of sector indices occupied by the qth aviation cable in the polar coordinate system, where u is the sector index. By using independent obstruction strength, the obstruction impact of each cable on the target cable can be intuitively quantified. If a cable is behind other obstruction cables (as shown in the figure, cable 2 is behind cable 10), the histogram value is 0, which means that the cable has no direct obstruction impact on the target cable.

[0033] S5, such as Figure 5 As shown, the topology segmentation path is pruned according to multi-level constraints including target avoidance constraints, capture quantity constraints, safety gap constraints, and target exposure angle constraints, retaining the topology segmentation path that satisfies the multi-level constraints; step S6 is executed; if there is no topology segmentation path that satisfies all four constraints after pruning, the process returns to step S1 and restarts after randomly adjusting the aviation cable bundle. In this embodiment, after constructing the topology segmentation path set in step S3, an optimal executable topology segmentation path needs to be obtained from it. To ensure that the final topology segmentation path is executable, paths that do not meet the grasping constraints need to be pruned. To ensure that the final topology segmentation path is optimal, the remaining executable paths need to be optimized. As a preferred implementation, the target avoidance constraints, grasping quantity constraints, safety gap constraints, and target exposure angle constraints in step S5 are specifically as follows: Target avoidance constraint: The target cable must be segmented into the reserved cluster rather than the grab cluster; Grab quantity constraint: The number of aviation cables in the grab cluster must not exceed the upper limit of the gripper capacity; Safety clearance constraint: The shortest distance from the center of all aviation cables to the topology split path must not be less than the minimum safety distance. The minimum safety distance is determined based on the flexible cable radius and the cable radial manipulation factor, expressed by the formula as follows: ; Target exposure angle constraint: Constrains the angle of the continuous exposure sector obtained by the target cable. It must not be less than the preset angle threshold. ( (Generally, 90°~120° is used); the angle of the continuously exposed sector is the maximum value of the sum of the central angles of the continuous sectors in the polar coordinate system after removing all obstruction cables in the capture cluster, where the obstruction histogram value is always zero, as shown in the formula: ; satisfy , ;in, The histogram values ​​are regenerated after removing the grab cluster cables; This represents the m-th consecutive interval of zero values ​​in the histogram.

[0034] S6. Considering optimization indicators including the number of captures, the difficulty of safe traversal, and the impact of occlusion, perform multi-objective optimization decision-making on the topology segmentation path to obtain the optimal segmentation path; In a preferred embodiment, the occlusion effect is determined by the cumulative occlusion intensity of the obstacle cables in the grasping cluster. ,in, To capture the cable cluster and the continuous exposed sector angle of the target cable. composition; The safety crossing difficulty is designed to quantify the spatial obstacle avoidance risk of a robot moving along a segmented path. It is composed of the number of cables traversed and a spatial spacing factor, specifically: First, calculate the shortest distance from the center of any aviation cable connector to the topology split path. If satisfied If so, the cable is determined to be a path crossing cable; in, The connector radius of the aviation cable. Find the minimum radius among all cable connector radii; iterate through all aviation cables to obtain the number of cables the path crosses; Spacing factor Calculated using the following formula: ; in, , These represent the minimum and average distances between the capture safety threshold circle for all paths traversing cables and the topology segmentation path, respectively. and These are the weights of the corresponding values, and their sum is 1.

[0035] In a preferred embodiment, step S6 specifically includes: The number of samples captured, the impact of occlusion, and the difficulty of safe passage are normalized; the normalization formula is: ,in This represents data on various influencing factors, such as the number of data captured and the angle of continuous gaps. and These represent the minimum and maximum values ​​of the corresponding influence factor data in all combinations of topological segmentation paths; After normalization, a global objective function is established based on a linear weighted average of three factors: the number of captures, the impact of occlusion, and the difficulty of safe traversal. Its formula is expressed as: ; in, These are the weighting factors corresponding to each optimization objective; , , , and The normalized values ​​are: number of captured items, number of traversed cables, traversed space spacing factor, cumulative occlusion intensity of obstacle cables, and angle of continuous gaps after removal of captured clusters.

[0036] The topological segmentation path with the largest global objective function is selected as the optimal segmentation path for final execution; for example... Figure 6 As shown in the figure, the blue solid curve represents the final optimal topology segmentation path, which divides the cable into grab clusters (green) and retain clusters (purple), without intruding into the grab safety threshold circles (yellow) on both sides. The corresponding histogram below shows that after removing the grab clusters, a continuous 120° unobstructed gap is formed in the 160°~280° range, satisfying the active decoupling grab condition.

[0037] S7. The dual-arm robot uses an active decoupling method to collaboratively complete the grasping operation based on the optimal segmentation path; As a preferred embodiment, such as Figure 7 As shown, step S7 specifically includes: A dual-arm robot consisting of a gripping arm and a decoupling arm performs the gripping task. First, the decoupling arm moves along an optimal segmentation path to gather, clamp, and remove the non-target aviation cable. The removal displacement satisfies the cable's flexible movement range constraint. This constraint states that the removal displacement of the obstacle cable must not exceed the preset movable range of the aviation cable. Because the aviation cable is suspended (i.e., the other end is fixed), and the cable is not stretchable, if the displacement exceeds the length of the suspended portion (i.e., the preset movable range), excessive pulling will damage the aviation cable. Therefore, this application imposes a cable flexible movement range constraint. The gripper uses a vision system to acquire the pose of the exposed target cable, plans a trajectory to move to the target point and close the gripper to complete precise gripping. According to the requirements of subsequent tasks, perform cable collaborative handover; the gripping arm and the decoupling arm cooperate in the collaborative area to transfer the target cable from the former to the end gripper of the latter.

[0038] In summary, the topology-segmentation-based robotic active decoupled aerospace cable grasping method proposed in this invention constructs a geometric analysis model of the aerospace cable cross-section to plan all possible topology-segmentation paths for the aerospace cable bundle. Then, path pruning is performed based on multi-level constraints such as target avoidance, grasping quantity, safety gap, and target exposure angle. Using grasping quantity, safe crossing difficulty, and occlusion impact as optimization indicators, the optimal segmentation path is obtained through multi-objective optimization decision-making. This enables a dual-arm robot to actively and decoupledly grasp the target cable based on the optimal segmentation path. This invention achieves reliable grasping of targets under dense occlusion of aerospace cables, providing an effective solution for the dexterous manipulation of cables during robotic cable assembly, thereby promoting the development of automated cable assembly technology in the aerospace field.

[0039] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0040] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0041] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A robot-based active decoupled aviation cable grasping method based on topology segmentation, characterized in that, Specifically, the following steps are included: S1. Obtain the position and radius information of each aviation cable in the suspended aviation cable bundle; the aviation cable is composed of flexible cable and cable connector; S2. Set a capture safety threshold, construct a cross-sectional geometric analysis model of the aviation cable based on its location and radius information, and label the target cable and add obstacle constraints; specifically: Calculate the horizontal cross-sectional height suitable for the robot's active decoupling operation, the horizontal cross-sectional height It is obtained by adding the maximum height of the cable connector center, half of the maximum length of the cable connector, the maximum length of the cable connector tail accessory, and the preset cross-sectional safety margin; Constructing a cross-sectional geometric analysis model based on top-view projection: taking Using the plane as the analysis section, the spatial distribution of each aviation cable is projected onto the analysis section. Based on the horizontal and vertical coordinates of the center of each aviation cable, concentric circles with the radius of the cable connector, the radius of the flexible cable, and the gripping safety threshold are drawn on the analysis section, respectively, and are denoted as the cable connector circle, the flexible cable circle, and the gripping safety threshold circle. These concentric circles are used as simplified projections of each aviation cable on the analysis section. Record the serial number of each aviation cable and mark the target cable; at the same time, based on the positional relationship between external obstacles and aviation cable bundles, simplify external obstacles into straight line boundaries on the cross section, as obstacle constraints for subsequent topology analysis; S3. Based on the position and geometric relationship between each aviation cable in the cross-sectional geometric analysis model, construct a topology segmentation path set; specifically including: S31. In the cross-sectional geometric analysis model, select any two aviation cables. Using the grab safety threshold circles of these two aviation cables as a reference, construct a combined ellipse whose major axis is collinear with the line connecting the centers of the two grab safety threshold circles and whose foci are the centers of the two grab safety threshold circles. The combined ellipse satisfies the constraint of including the grab safety threshold circles and the boundary safety margin. The semi-major axis of the combined ellipse is determined by the distance between the centers of the two grab safety threshold circles and the larger radius of the two grab safety threshold circles. The semi-minor axis of the combined ellipse is determined by the semi-major axis and the preset shape adjustment ratio constant. The constraint of including the grab safety threshold circles and the boundary safety margin is that the combined ellipse must completely include the two grab safety threshold circles, and the shortest distance from any point on the combined ellipse to the two grab safety threshold circles is not less than the preset safety margin. S32. Draw perpendicular lines to the major axis of the combined ellipse through the centers of the flexible cables of the two aviation cables respectively. Take the intersection of the perpendicular line and the combined ellipse as the tangent point. Draw tangent lines to the combined ellipse based on the tangent point. The continuous path that connects the tangent segment, the ellipse arc segment, and the tangent segment in sequence is the topology segmentation path. The topology segmentation path divides the analysis section into a grab cluster located inside the topology segmentation path and a retain cluster located outside the topology segmentation path. Each combined ellipse generates two topology segmentation paths. S33. Repeat steps S31-S32 for all aviation cable combinations to generate a topology partitioning path set containing all potential partitioning schemes. S4. Based on the vector field histogram algorithm, quantify the independent occlusion intensity of each obstacle cable on the target cable; S5. Prune the topology segmentation path according to multi-level constraints including target avoidance constraints, capture quantity constraints, safety gap constraints, and target exposure angle constraints, and retain the topology segmentation path that satisfies the multi-level constraints; execute step S6; if there is no topology segmentation path that satisfies all four constraints after pruning, return to step S1 and start execution again after randomly moving the aviation cable bundle. S6. Considering optimization indicators including the number of captures, the difficulty of safe traversal, and the impact of occlusion, perform multi-objective optimization decision-making on the topology segmentation path to obtain the optimal segmentation path; S7. The dual-arm robot uses an active decoupling method to collaboratively complete the grasping operation based on the optimal segmentation path.

2. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, The gripping safety threshold mentioned in step S2 is composed of the radius of the aviation flexible cable, the product of the cable radial manipulation factor and the radius of the aviation flexible cable, the half width of a single clamping piece of the gripper, the system error margin, and the preset safety margin; the cable radial manipulation factor is determined based on the flexible deformation characteristics of the aviation cable.

3. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, Step S4 specifically includes: A polar coordinate system is constructed with the center of the flexible cable circle of the target cable as the origin, and the analysis section is divided into N sectors, each sector containing K virtual ray network units; the obstacle strength of each virtual ray network unit is calculated based on preset deterministic values, preset constant parameters, and the nearest Euclidean distance from the ray origin to the obstacle. The obstacle strengths of all virtual ray network units in each sector are summed to obtain the obstacle histogram value for each sector; For an obstruction cable, calculate the sum of the obstruction histogram values ​​of all sectors it occupies, which is taken as the independent blocking strength of the obstruction cable to the target cable.

4. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, The target avoidance constraints, capture quantity constraints, safety gap constraints, and target exposure angle constraints in step S5 are as follows: Target avoidance constraint: The target cable must be segmented into the reserved cluster rather than the grab cluster; Grab quantity constraint: The number of aviation cables in the grab cluster must not exceed the upper limit of the gripper capacity; Safety clearance constraint: The shortest distance from the center of all aviation cables to the topology split path must not be less than the minimum safety distance, which is determined based on the flexible cable radius and the cable radial manipulation factor; Target exposure angle constraint: The angle of the continuous exposed sector obtained by the target cable shall not be less than the preset angle threshold; the angle of the continuous exposed sector is the maximum value of the sum of the central angles of the continuous sectors with zero obstacle histogram value in the polar coordinate system after removing all obstacle cables in the grab cluster.

5. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 4, characterized in that, In step S6, the occlusion effect consists of the cumulative occlusion intensity of the obstacle cables in the capture cluster and the continuous exposure sector angle of the target cable.

6. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, In step S6, the safety crossing difficulty is composed of the number of cables traversed along the path and the spatial spacing factor, specifically: First, calculate the shortest distance from the center of any aviation cable connector to the topology split path. If satisfied If so, the cable is determined to be a path crossing cable; in, The connector radius of the aviation cable. Find the minimum radius among all cable connector radii; iterate through all aviation cables to obtain the number of cables the path crosses; Spatial separation factor is calculated from the following calculation formula: ； in, , These represent the minimum and average distances between the capture safety threshold circle for all paths traversing cables and the topology segmentation path, respectively. and These are the weights of the corresponding values, and their sum is 1.

7. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, Step S6 is as follows: The number of samples captured, the impact of occlusion, and the difficulty of safe passage were normalized. After normalization, a global objective function is established based on a linear weighted average of three factors: the number of captures, the impact of occlusion, and the difficulty of safe passage. The topological segmentation path with the largest global objective function is selected as the optimal segmentation path for final execution.

8. The robot-based active decoupled aviation cable grasping method based on topology segmentation according to claim 1, characterized in that, Step S7 specifically includes: A dual-arm robot consisting of a gripping arm and a decoupling arm is used to perform the gripping task. First, the decoupling arm moves along the optimal segmentation path to gather, clamp, and remove the non-target aviation cable. The removal displacement meets the cable's flexible range of motion constraint. The cable's flexible range of motion constraint is that the removal displacement of the obstacle cable must not exceed the preset movable range of the aviation cable. The gripper uses a vision system to acquire the pose of the exposed target cable, plans a trajectory to move to the target point and close the gripper to complete precise gripping. According to the requirements of subsequent tasks, perform cable collaborative handover; the gripping arm and the decoupling arm cooperate in the collaborative area to transfer the target cable from the former to the end gripper of the latter.

Images