Method, system and medium for adaptive planning of automated scanning paths for cabin structures

Abstract

The application provides a kind of cabin structure automatic scanning path adaptive planning method, system and medium, comprising: in the preliminary scanning stage, control scanner along preset initial path scans the inner surface of cabin, based on the depth information of the nearest point in the current frame point cloud obtained in real time from the inner surface, collision detection is carried out and the pose of scanner is dynamically adjusted in response to collision risk to generate collision-free preliminary scanning path;Based on the registration of preliminary scanning point cloud and theoretical numerical model, the spatial conversion relationship is obtained, and the support centroid position is mapped to the actual physical space;With the support centroid after mapping as the center of the sphere, a spherical scanning point set is generated, and a detailed scanning path is planned accordingly to control the scanner to scan the support structure.The application realizes online adaptive obstacle avoidance without additional sensors and detailed scanning of complex support structure.

Description

Technical Field

[0001] This invention relates to the field of automated measurement technology, and more specifically, to an adaptive planning method, system, and medium for automated scanning path planning of cabin structures. Background Technology

[0002] In the automated measurement of large cabin / section components, scanning path planning faces enormous challenges due to the characteristics of these components, such as large size, continuous curved surfaces and complex occlusion, and limited space for collaborative movement of on-site tooling and robots. Once the scanning path is generated offline in the automated measurement process, collision risks often occur during actual execution due to environmental differences or local interference, which requires repeated manual adjustments to the position and path, resulting in low efficiency and difficulty in standardization.

[0003] In the prior art, Chinese patent document with publication number CN114719775B proposed an automated scanning path planning and regional measurement and splicing method for inner and outer surfaces based on a 3D CAD model. However, it focuses on shape reconstruction and splicing accuracy and does not establish an online adaptive obstacle avoidance mechanism for the initial path collision area.

[0004] In Chinese patent document CN120101689A, a method was proposed to divide the scanning area according to the field of view of the scanning device to improve efficiency. However, this method is biased towards offline or semi-offline planning paradigms and does not adequately cover the online replanning of collision areas during the robot-turntable collaborative execution process. In Chinese patent document CN112577447A, a fully automatic scanning path planning method based on CAD triangulation viewpoint sampling is disclosed. However, its alignment process requires manual selection of corresponding points and relies on marker points for assistance, and it does not form an online closed loop of "collision occurrence - real-time obstacle detection - dynamic path adjustment". Chinese patent document CN112325796A discloses a multi-view point cloud stitching reconstruction method with emphasis on large scene positioning assistance, but does not involve the adaptive dynamic obstacle avoidance and hierarchical planning mechanism of the scanning path during execution.

[0005] In summary, existing solutions generally suffer from problems such as the ease with which offline fixed paths can lead to collisions due to environmental differences during actual execution, high system redundancy costs due to reliance on additional external sensors, and insufficient scanning precision for complex internal support structures. These issues make it difficult to simultaneously meet the integrated requirements of online adaptive planning of collision areas and precise scanning driven by key poses in automated measurement of large cabins. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the purpose of this invention is to provide an adaptive planning method, system, and medium for automated scanning paths of cabin structures.

[0007] An adaptive planning method for automated scanning path of a cabin structure, provided by the present invention, includes: The scanner is controlled to scan the inner surface of the cabin structure along a preset initial path, and collision detection is performed based on the depth information of the point closest to the inner surface in the current frame point cloud acquired by the scanner in real time during the scanning process. In response to the detection of a collision risk, the pose of the scanner is dynamically adjusted radially to generate a collision-free initial scanning path; The preliminary scan point cloud obtained based on the preliminary scan path is registered with the theoretical digital model of the cabin structure to obtain the spatial transformation relationship, and the position of the support centroid in the theoretical digital model is mapped to the actual physical space based on the spatial transformation relationship. A spherical scanning point set is generated with the mapped centroid of the support as the center of the sphere, and a refined scanning path is planned based on the spherical scanning point set. The scanner is then controlled to scan the support structure along the refined scanning path.

[0008] Preferably, the radial dynamic adjustment of the scanner's pose in response to the detection of a collision risk includes: The point closest to the inner surface of the cabin structure in the current frame point cloud is obtained as the depth information; the scanner is controlled to move radially closer to or further away from the inner surface of the cabin structure based on the depth information.

[0009] Preferably, controlling the scanner to move radially closer to or further away from the inner surface of the cabin structure based on the depth information includes: Set the maximum distance threshold and the minimum distance threshold; When the depth information is greater than the maximum distance threshold, the scanner is controlled to move radially closer to the inner surface of the cabin structure. When the depth information is less than the minimum distance threshold, the scanner is controlled to move radially away from the inner surface of the cabin structure. When a collision risk is detected, the scanner's orientation is adjusted so that its field of view faces the direction of approach from the inner surface of the cabin structure.

[0010] Preferably, generating a spherical scan point set with the mapped centroid of the scaffold as the sphere center includes: Using the mapped centroid of the support as the center of a sphere and the preset optimal scanning distance as the radius, the spherical scanning point set is generated; The process of planning a refined scanning path based on the spherical scan point set includes: For the path portion where collisions occur in the spherical scanning point set, a fast obstacle-free path planning algorithm is used to generate an avoidance path, thus obtaining the refined scanning path.

[0011] Preferably, generating the spherical scan point set includes: Using the centroid of the support as the center, spatial circles are generated at preset angles along the X-axis, Y-axis, and Z-axis respectively. The intersections between the spatial circles are taken as multiple uniformly distributed spherical scanning points in the spherical scanning point set. The normal of each spherical scanning point is the line connecting itself to the center of the sphere. The fast, barrier-free path planning algorithm is a fast planning method from the RRT series.

[0012] Preferably, controlling the scanner to scan the support structure along the refined scanning path includes: The spherical scan point set is arranged in descending order of Z value; For each of the stent centroids, the following actions are performed sequentially: Control the turntable to rotate to the motion angle corresponding to the projection of the support's center of mass onto the turntable plane; Control the three-degree-of-freedom moving mechanism to move to the partition height corresponding to the center of mass of the support; The six-degree-of-freedom robot is controlled to perform linear motion on the spherical scanning points according to the generation order of the spherical scanning points in the set of spherical scanning points.

[0013] Preferably, before the controlled scanner scans the inner surface of the cabin structure along a preset initial path, the method further includes: The cabin structure is longitudinally partitioned based on the Z-axis range of the robot's reachable space, resulting in multiple partitions; Calculate the coordinate points of the three-degree-of-freedom moving mechanism corresponding to each of the partition boundaries; Based on the theoretical shape of the cabin structure and the field of view of the scanner, the preset initial path is generated.

[0014] Preferably, before longitudinally partitioning the cabin structure within the Z-axis range of the robot-based reachable space, the method further includes: The position of the target ball is acquired using a laser tracker. A coordinate system for the turntable and a coordinate system for the three-degree-of-freedom moving mechanism are established by fitting the cylindrical surface and the plane, so as to obtain the calibration relationship between the coordinate system for the turntable and the coordinate system for the three-degree-of-freedom moving mechanism.

[0015] An automated scanning path adaptive planning system for a cabin structure, according to the present invention, includes: The preliminary scan planning module is used to control the scanner to scan the inner surface of the cabin structure along a preset initial path during the preliminary scan stage. During the scan, collision detection is performed based on the depth information of the point closest to the inner surface in the current frame point cloud acquired by the scanner in real time. In response to the detection of collision risk, the pose of the scanner is dynamically adjusted radially to generate a collision-free preliminary scan path. The feature mapping module is used to register the preliminary scan point cloud obtained based on the preliminary scan path with the theoretical digital model of the cabin structure, obtain the spatial transformation relationship, and map the support centroid position in the theoretical digital model to the actual physical space based on the spatial transformation relationship. The refined scanning planning module is used to generate a spherical scanning point set with the mapped centroid of the support as the center of the sphere, and to plan a refined scanning path based on the spherical scanning point set, and control the scanner to scan the support structure along the refined scanning path.

[0016] According to the present invention, a computer-readable storage medium is provided thereon storing a computer program, which, when executed by a processor, implements the aforementioned adaptive planning method for automated scanning path of cabin structure.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention utilizes the depth information of the nearest point to the inner surface in the current frame point cloud acquired in real time by the scanner itself during the preliminary scanning stage to perform collision detection, and performs radial dynamic adjustment of the scanner pose in response to collision risk. This enables online safe path exploration and flexible obstacle avoidance without adding additional external sensors, effectively solving the offline path collision problem caused by manufacturing errors or environmental differences in large cabins.

[0018] 2. This invention obtains the spatial transformation relationship based on the registration of the preliminary scan point cloud and the theoretical digital model, accurately maps the centroid position of the support in the theoretical digital model to the actual physical space, and generates a spherical scan point set with the centroid as the center to plan a refined scan path. This achieves adaptive alignment between theoretical design features and actual physical positions, overcomes the limitations of traditional methods that rely on manual point selection or global marker points, and ensures a high-precision full-coverage scan with no blind spots for complex support structures inside the cabin.

[0019] 3. This invention divides the scanning process into two stages: preliminary scanning and fine scanning. In the fine scanning stage, a fast and unobstructed path planning algorithm and a multi-device collaborative motion strategy are combined to achieve a seamless connection from the safe and rapid acquisition of macroscopic contours to the fine detection of microscopic features. While ensuring the absolute safety of the measurement process, it significantly improves the overall efficiency and standardization level of automated measurement of large cabin structures. Attached Figure Description

[0020] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a partial flowchart of the adaptive flexible dynamic programming method for scanning paths according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the composition structure of the large cabin automated scanning system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of cabin partitioning and initial path planning based on the robot's reachability according to an embodiment of the present invention; Figure 4 This is a flowchart of the radial adjustment process of the robot end effector based on the Z-value determination of point cloud according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the spherical scanning point distribution structure around the centroid of this invention.

[0021] Explanation of reference numerals in the attached drawings: 1-Large field-of-view binocular measuring equipment, 2-Three-degree-of-freedom motion mechanism, 3-Six-degree-of-freedom robot, 4-Scanner, 5-Large cabin product, 6-Turntable. Detailed Implementation

[0022] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0023] This embodiment provides an adaptive scanning path planning method for automated measurement of large cabin structures, which is applied to an automated scanning system for large cabin structures.

[0024] like Figure 2As shown, the hardware configuration of the system includes a large cabin product 5, a turntable 6, a six-degree-of-freedom robot 3, a positioning fixture, a three-degree-of-freedom motion mechanism 2, a scanner 4, and a large field-of-view binocular measuring device 1. The 3D scanner 4 is mounted on the end flange of the six-degree-of-freedom robot 3. The large cabin product 5 is mounted on the turntable 6 via the positioning fixture. The base of the six-degree-of-freedom robot 3 is fixedly mounted on the three-degree-of-freedom motion mechanism 2. The large field-of-view binocular measuring device 1 is mounted on the top beam of the system, and its field of view covers the entire movement range of the turntable 6 and the scanner 4. The turntable 6 has etched lines on its surface. During installation, the quadrant lines of the large cabin product 5 are aligned with the etched lines on the turntable 6 to ensure the accuracy of the product's installation orientation. Scattered target points are affixed to the upper surface of the turntable 6. The large field-of-view binocular measuring device 1 continuously observes these scattered target points and the scanner 4 throughout the scanning process to provide global positioning information. The cabin product has a cylindrical or conical structure, and its inner surface is fitted with irregular support structures. These support structures are key targets for fine-tuning the scanning process.

[0025] Before the system runs, three sets of coordinate systems need to be defined uniformly: a turntable coordinate system with the intersection of the cylindrical axis of the turntable 6 and the upper plane as the origin; a three-degree-of-freedom moving mechanism coordinate system based on the motion axis of the three-degree-of-freedom moving mechanism 2; and a target point coordinate system based on the scattered target points on the upper surface of the turntable 6. The method in this embodiment specifically includes the following steps.

[0026] Step S1: Establish the hardware configuration for the automated scanning system of the large cabin structure.

[0027] Step S2: Due to errors in the system hardware assembly, to ensure more accurate subsequent path planning, a laser tracker is used to calibrate the relative relationship between the turntable coordinate system and the three-degree-of-freedom moving mechanism coordinate system. In addition, the transformation relationship between the target point coordinate system on the turntable 6 and the turntable coordinate system also needs to be calibrated.

[0028] Specifically, it includes the following sub-steps: Step S21: Calibrate the turntable 6 using a laser tracker. With the target ball of the laser tracker pressed against the cylindrical surface of the turntable 6, continuously acquire and move the position of the target ball to obtain several points on the cylindrical surface of the turntable 6. Fit the cylindrical surface to obtain the axis of the cylindrical surface of the turntable 6. Press the target ball against the upper plane of the turntable 6, continuously change the measurement position and record the position data. Obtain the upper plane of the turntable 6 by fitting the plane and offsetting by one target ball radius. Use the intersection point of the obtained axis of the cylindrical surface of the turntable 6 and the upper plane of the turntable 6 as the origin of the turntable coordinate system, and the axis of the cylindrical surface of the turntable 6 as the Z-axis. Place the target ball close to the 0-degree mark on the upper plane of the turntable 6, measure the point, and connect the point to the origin as the X-axis, thus establishing the turntable coordinate system.

[0029] Step S22: Keep the laser tracker position unchanged and the three-degree-of-freedom moving mechanism 2 stationary. Record the position values ​​of the three main axes of the moving mechanism at this time. The target ball is placed on the end mounting plane of the three-degree-of-freedom moving mechanism 2. The measurement position is continuously changed while keeping the target ball close to the plane to obtain several measurement points on the end mounting plane. The mounting plane is then fitted. The target ball is then pressed against the cylindrical surface of the third main shaft of the moving platform. The measurement position is continuously changed while keeping the target ball close to the platform to obtain several points on the cylindrical surface of the main shaft. The cylindrical axis is then fitted to the cylindrical surface. There is an intersection point between the cylindrical axis and the fitted mounting plane. .

[0030] Step S23: Based on step S22, another target ball is randomly attached to the mounting plane at the end of the three-degree-of-freedom moving mechanism 2. The first main axis of the moving platform, i.e. the X-axis, is driven and real-time measurement is maintained to obtain several measurement points in the X direction. A straight line in the X direction is obtained through multi-point fitting.

[0031] Step S24: Establish the coordinate system of the three-degree-of-freedom translating mechanism. The origin is... The straight line obtained in step S23 is taken as the X-axis, and the cylindrical axis obtained in step S22 is taken as the Z-axis, thus obtaining the coordinate system of the three-degree-of-freedom moving mechanism.

[0032] Steps S25, S21, and S24 respectively obtain the turntable coordinate system and the three-degree-of-freedom moving mechanism coordinate system under the laser tracker coordinate system, and the calibration relationship between the turntable coordinate system and the three-degree-of-freedom moving mechanism coordinate system can be obtained. This calibration process provides a precise physical space transformation basis for all subsequent path planning and eliminates the impact of hardware assembly errors on scanning accuracy.

[0033] Step S26: Place the laser tracker target ball sequentially at the attachment position of the planar visual target point on the turntable 6, measure and record the position of the target ball coordinates in the turntable coordinate system, and translate it in the negative Z-axis direction of the turntable 6 by a distance equal to the radius of the target ball to obtain the position of the visual target point in the turntable coordinate system.

[0034] Step S27: Based on the coordinate system establishment rules of the large field-of-view scanner 4 for the target points scattered on the plane of the turntable 6, calculate the transformation relationship between the turntable coordinate system and the target point coordinate system.

[0035] Step S3: Based on step S2, establish a large cabin structure automated scanning system model in system simulation and visualization software.

[0036] Specifically, the digital models of equipment and products in the automated scanning system for large cabin structures are imported into the software. Based on the calibration relationship between the turntable coordinate system and the three-degree-of-freedom mobile mechanism coordinate system obtained in step S2, the kinematic pairs and forward and inverse kinematic solutions of the robot, turntable 6 and three-degree-of-freedom mobile mechanism 2 are established. The motion of the robot, turntable 6 and three-degree-of-freedom mobile mechanism 2 is mapped in real time through communication protocols such as TCP and Modbus to complete the virtual mapping of the overall scanning system, with the three-degree-of-freedom mobile mechanism coordinate system as the world coordinate system.

[0037] Step S4, refer to Figure 1 As shown, using the Creo secondary development plugin, the key dimensions of the cabin product are exported, including the theoretical height of the cabin. The key dimensional data includes the diameters of the large and small ends of the hull, and the position of the centroid of the supports mounted on the hull surface relative to the digital model's world coordinate system. These critical dimensional data serve as the foundational inputs for subsequent path planning and the generation of the spherical scan point set.

[0038] Step S5: Based on the robot's reachable space and the theoretical shape of the cabin product, the system simulation and visualization software automatically plans several Z-direction height values ​​for the three-degree-of-freedom mobile mechanism 2.

[0039] like Figure 3 As shown, the specific steps include the following: Step S51: Calculate the reachability space of the selected robot. Use the Monte Carlo method to randomly sample the robot's overall joint space. Map each sampled point to three-dimensional space using forward kinematics to form a reachability point cloud. Find the points with the maximum and minimum Z-values ​​to obtain the Z-axis range of the reachability space. .

[0040] Step S52, based on the Z-axis range of the reachable space obtained in step S51. The theoretical height of the cabin obtained in step S4 The cabin is divided into longitudinal sections, each with a height of [missing information]. It is divided into 100 parts. Each partition is defined as follows. Since step S2 has already established the relationship between the turntable coordinate system and the three-degree-of-freedom (DOF) traversal mechanism coordinate system, the height values ​​of each partition boundary in the three-DOF traversal mechanism coordinate system can be obtained. Because the XY coordinates of the robot's base coordinate system origin are located on the central axis of the cabin product and can cover the diameter processing range of the cabin product, the coordinates of each partition's robot base coordinate system in the world coordinate system can be obtained. ,like Figure 3 As shown, the three-degree-of-freedom moving mechanism 2 is then calculated. When scanning the product surface in a certain zone, the three-degree-of-freedom moving mechanism 2 moves to the corresponding coordinate point. This automatic zoning strategy based on reachable space ensures that the robot can cover the maximum area in a suitable posture at each station, avoiding efficiency losses caused by frequent station changes.

[0041] Step S6: Based on the cylindrical or conical theoretical shape of the cabin product and the field of view of the scanner 4, the system simulation and visualization software automatically plans the robot's initial path.

[0042] Specifically, it includes the following sub-steps: In step S61, during the scanning process, the cabin continuously performs circular motion driven by the turntable 6. The movement trajectory of the robot end-effector scanner 4 is parallel to the generatrix of the cabin's conical surface and is at a fixed scanning distance from the generatrix. Without considering collisions, the robot end-effector scanner 4 slowly moves downwards parallel to the generatrix to complete the rough scan of the entire inner surface of the cabin. Figure 3 As shown.

[0043] Step S62, based on the coarse scanning strategy of step S61 and the theoretical height and diameter of the large and small ends of the cabin obtained in step S4, can calculate the spatial linear motion trajectory equation of the robot in the coordinate system of the three-degree-of-freedom mobile mechanism. Substituting the Z-values ​​of the partition boundaries from step S52 into the spatial linear motion equation, the sampling points of the robot's motion trajectory at each partition boundary are calculated, such as... Figure 3 As shown.

[0044] Step S63: The robot's base coordinate system has a different height in each partition. Step S62: Several trajectory sampling points are given in the world coordinate system. Step S52: The coordinates of the robot's base coordinate system origin corresponding to each partition in the world coordinate system are given.

[0045] At this point, the robot's base coordinate system position for each partition is determined. Using each partition plane as a boundary, the motion trajectory from step S62 is divided, and the robot executes the corresponding motion trajectory for each partition, as follows: Figure 3 As shown.

[0046] In step S7, the system simulation and visualization software automatically simulates the initial path. If there is no collision interference during the process, the coarse scan path planning and obstacle avoidance part ends, and the fine scan path planning and obstacle avoidance stage begins. If a collision interference occurs during the process, the scanning device is used as a sensor to monitor obstacles in real time and dynamically adjust them radially to complete the coarse scan path planning and obstacle avoidance.

[0047] Specifically, it includes the following sub-steps: In step S71, based on step S6, before the robot and turntable 6 and other equipment move, the path planning in step S63 is simulated. If no collision interference occurs during the simulation, the coarse positioning scan path planning and obstacle avoidance part is completed, and the robot, turntable 6 and three-degree-of-freedom mobile mechanism 2 can be driven to move according to the initial plan.

[0048] Step S72: Based on step S71, if the robot's end-effector scanner 4 collides with the cabin structure during the simulation, it proves that the initial path is not feasible. The robot's end-effector posture is adjusted so that the scanner 4's field of view faces the turntable 6 towards the bottom surface of the cabin, enabling it to acquire the point cloud of the upcoming inner surface of the cabin. Figure 4 As shown.

[0049] Step S73, as follows Figure 4 As shown, after the scanner 4 acquires the point cloud of the current frame, it finds the point with the smallest Z value in the point cloud. Perform a search and set the maximum Z-value range. and the range of minimum Z values ,when The Z value of the point is greater than When this occurs, it indicates that scanner 4 is too far from the surface of the cabin; when The Z value of the point is less than When the scanner 4 is too close to the surface of the cabin, there is a risk of collision. Therefore, the end effector of the robot is driven radially away from the surface of the cabin.

[0050] This solution utilizes the scanner 4's own sensing data for closed-loop feedback, eliminating the need for additional LiDAR or ultrasonic sensors. This reduces system hardware costs and complexity, and avoids new error sources introduced by multi-sensor calibration, making it particularly suitable for space-constrained and environmentally variable scenarios such as the interior of large cabins. It should be understood that in this embodiment, depth information is specifically represented by the Z-value along the optical axis of the scanner 4 in the point cloud data. However, in other embodiments, depth information can also be the Euclidean distance from the point cloud to the center of the scanner 4 or the projected distance along the normal direction, as long as it can characterize the relative distance between the scanner 4 and the workpiece surface.

[0051] Step S8: After completing the coarse scan, obtain the coarse scan point cloud and register it with the theoretical digital model point cloud to form a transformation matrix between the scan point cloud and the theoretical digital model point cloud.

[0052] Specifically, it includes the following sub-steps: In step S81, during the scanning process, the scanner 4 can not only obtain the position information of the current frame product point cloud relative to the scanner 4, but also transfer all point clouds to the target point coordinate system through the target point on the upper surface of the turntable 6 to obtain a complete coarse scan point cloud.

[0053] Step S82 involves registering the coarse scan point cloud with the actual theoretical digital model point cloud. After coarse and fine registration, a transformation matrix between the coarse scan point cloud and the actual theoretical digital model point cloud is obtained. Using this transformation matrix, the pre-exported centroid coordinates of the support in the CAD model can be accurately converted to the current scanning coordinate system, thereby solving the problem of inconsistency between theoretical features and actual positions caused by workpiece manufacturing deviations or clamping deformation.

[0054] Step S9: Based on the transformation matrix obtained in step S8, the position of the support centroid in the digital model world coordinate system is transformed to the scanning point cloud. With the centroid as the center of the sphere and the optimal scanning distance of the scanner 4 as the radius, a spherical scanning point set is generated.

[0055] Specifically, it includes the following sub-steps: Step S91: Based on the transformation matrix obtained in step S82, the centroid position of the support derived in step S4 is transformed from the world coordinate system of the digital model to the target point coordinate system of the scanned point cloud.

[0056] Step S92, for each scaffold centroid At the optimal scanning distance Using a radius of 1, and with its own position as the center, generate 26 evenly distributed spherical scanning points. The normal of each scanning point is the line connecting itself to the center of the sphere. For example... Figure 5 As shown, the generation strategy is as follows: Using the centroid as the center, generate circles every 45 degrees along the X-axis, Y-axis, and Z-axis. A number of spatial circles are obtained with a radius of 26, and the intersection of each spatial circle is used as the center scanning point. It should be noted that the "26 points" and "45-degree interval" are merely specific examples for achieving uniform encirclement sampling and are not intended to limit the scope of protection of this invention. In practical applications, other regular polyhedral vertex distributions or uniform sampling strategies based on Fibonacci spherical spirals can also be used, as long as these point sets can form an enclosure around the support structure with the centroid as the center and meet the optimal working distance requirements of scanner 4, they all fall within the scope of protection of this invention. This geometrically feature-based encirclement sampling strategy, compared to traditional grid scanning or single-layer surround scanning, can more comprehensively capture the complex morphology of the sides, underside, and top of the support structure, improving reconstruction integrity.

[0057] Step S10: Automatically plan the position of turntable 6 and the initial path of the robot, and perform collision simulation on the formed initial path.

[0058] Specifically, it includes the following sub-steps: Step S101: For the set of spherical scanning points distributed at different positions on the inner surface of the cabin, the coordinates of the set of spherical scanning points are transformed from the target point coordinate system to the turntable coordinate system according to the transformation relationship in step S27, and then arranged in descending order according to the transformed Z value to obtain the set of spherical scanning points from top to bottom.

[0059] Step S102: Calculate the coordinate projection of the center of mass on the plane of the turntable 6, calculate the motion angle of the center of mass projection point when it moves to a uniform angle with the turntable 6, and obtain the motion angle of the turntable 6 corresponding to the center of mass.

[0060] Step S103: For the Z value of the spherical scanning point set, divide it according to the partitioning in step S6 to obtain the motion value of the three-degree-of-freedom moving mechanism 2 corresponding to each spherical scanning point.

[0061] Step S104, the planning rules are as follows: First, the spherical scanning point set is classified according to the centroid of the support, forming several subsets of spherical scanning points. Then, according to the negative Z direction of the turntable coordinate system, the robot sequentially executes the scanning path points of each subset. Specifically, for each centroid point from top to bottom, the turntable 6 first rotates according to step S102, then the three-degree-of-freedom moving mechanism 2 moves according to step S103, and finally the robot moves linearly to the 26 spherical scanning points according to the generation order of the spherical scanning points. After the movement is completed, the robot returns to the initial state, the three-degree-of-freedom moving mechanism 2 returns to the initial state, and the turntable 6 moves to the next movement angle, repeating the cycle until the end. This hierarchical collaborative strategy of "turntable 6 positioning - lifting mechanism height adjustment - robot fine scanning" makes full use of the large-range rotation capability of the turntable 6 and the vertical coverage capability of the moving mechanism, limiting the robot's movement to its optimal working range and avoiding shaking and accuracy loss caused by overextension or strange postures of the robot.

[0062] In step S105, the system software performs reachability collision detection on the spherical scanning points in step S104. If the spherical scanning point cannot be reached or the robot interferes with the product when it arrives, the corresponding spherical scanning point is removed to form a new set of spherical scanning points.

[0063] In step S106, the system software performs collision simulation on the initial path from step S105. If there is no interference, the system directly drives the real device to move according to this plan, and the fine scan ends.

[0064] Step S11: For the collision-prone path portion, a fast obstacle-free path planning algorithm is used to form a collision-free path. Specifically, if collision interference occurs in step S106, the RRT series of fast planning methods is used on the path between the spherical scanning points to provide a collision-free avoidance path, resulting in a refined scanning path. It should be understood that the RRT algorithm is only one implementation method; other algorithms with equivalent functionality, as long as they can generate collision-free connection paths based on environmental constraints, can be used as alternatives.

[0065] In step S12, based on step S9, the robot turntable 6 is moved to complete the fine scan, and the overall process ends. The coarse scan provides the approximate structural point cloud of the cabin structure, and the fine scan provides the refined scan point cloud of the support structure. By combining the two, complete high-precision three-dimensional data of the inner surface of the cabin structure and the support structure can be obtained.

[0066] Based on the same inventive concept, this embodiment also provides a scanning path adaptive planning device for automated measurement of large cabin structures, including: a preliminary scanning planning module, a feature mapping module, and a refined scanning planning module.

[0067] The preliminary scan planning module is used to control the scanner 4 to scan the inner surface of the cabin structure along a preset initial path during the preliminary scan stage. During the scan, collision detection is performed based on the depth information of the point closest to the inner surface in the current frame point cloud acquired by the scanner 4 in real time. In response to the detection of collision risk, the pose of the scanner 4 is dynamically adjusted radially to generate a collision-free preliminary scan path.

[0068] The feature mapping module is used to register the preliminary scan point cloud obtained based on the preliminary scan path with the theoretical digital model of the cabin structure, obtain the spatial transformation relationship, and map the position of the support centroid in the theoretical digital model to the actual physical space based on the spatial transformation relationship.

[0069] The refined scanning planning module is used to generate a spherical scanning point set with the mapped centroid of the support as the center of the sphere, and to plan a refined scanning path based on the spherical scanning point set, controlling the scanner 4 to scan the support structure along the refined scanning path. Each module in this device can be an independent unit implemented by software, hardware, or firmware, or it can be a functional component integrated in the same processor. Its specific execution logic is completely consistent with the aforementioned method embodiments, and will not be repeated here.

[0070] Furthermore, this embodiment also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps described in the above method embodiments. This storage medium can be a tangible non-volatile memory, such as ROM, EEPROM, flash memory, optical disk, or magnetic disk, or any other medium capable of carrying program code. By embedding the above method into a software product, it facilitates deployment and upgrades in different types of automated measurement systems.

[0071] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0072] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0073] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features of the present invention can be arbitrarily combined with each other.

Claims

1. An adaptive planning method for automated scanning path of a cabin structure, characterized in that, include: The scanner is controlled to scan the inner surface of the cabin structure along a preset initial path, and collision detection is performed based on the depth information of the point closest to the inner surface in the current frame point cloud acquired by the scanner in real time during the scanning process. In response to the detection of a collision risk, the pose of the scanner is dynamically adjusted radially to generate a collision-free initial scanning path; The preliminary scan point cloud obtained based on the preliminary scan path is registered with the theoretical digital model of the cabin structure to obtain the spatial transformation relationship, and the position of the support centroid in the theoretical digital model is mapped to the actual physical space based on the spatial transformation relationship. A spherical scanning point set is generated with the mapped centroid of the support as the center of the sphere, and a refined scanning path is planned based on the spherical scanning point set. The scanner is then controlled to scan the support structure along the refined scanning path.

2. The adaptive planning method for automated scanning path of cabin structure according to claim 1, characterized in that, The radial dynamic adjustment of the scanner's pose in response to the detection of a collision risk includes: The point closest to the inner surface of the cabin structure in the current frame point cloud is obtained as the depth information; the scanner is controlled to move radially closer to or further away from the inner surface of the cabin structure based on the depth information.

3. The adaptive planning method for automated scanning path of cabin structure according to claim 2, characterized in that, The step of controlling the scanner to move radially closer to or further away from the inner surface of the cabin structure based on the depth information includes: Set the maximum distance threshold and the minimum distance threshold; When the depth information is greater than the maximum distance threshold, the scanner is controlled to move radially closer to the inner surface of the cabin structure. When the depth information is less than the minimum distance threshold, the scanner is controlled to move radially away from the inner surface of the cabin structure. When a collision risk is detected, the scanner's orientation is adjusted so that its field of view faces the direction of the approach from the inner surface of the cabin structure.

4. The adaptive planning method for automated scanning path of cabin structure according to claim 1, characterized in that, The step of generating a spherical scanning point set with the mapped centroid of the stent as the sphere center includes: Using the mapped centroid of the support as the center of a sphere and the preset optimal scanning distance as the radius, the spherical scanning point set is generated; The process of planning a refined scanning path based on the spherical scan point set includes: For the path portion where collisions occur in the spherical scanning point set, a fast obstacle-free path planning algorithm is used to generate an avoidance path, thus obtaining the refined scanning path.

5. The adaptive planning method for automated scanning path of cabin structure according to claim 4, characterized in that, The generation of the spherical scan point set includes: Using the centroid of the support as the center, spatial circles are generated at preset angles along the X-axis, Y-axis, and Z-axis respectively. The intersections between the spatial circles are taken as multiple uniformly distributed spherical scanning points in the spherical scanning point set. The normal of each spherical scanning point is the line connecting itself to the center of the sphere. The fast, barrier-free path planning algorithm is a fast planning method from the RRT series.

6. The adaptive planning method for automated scanning path of cabin structure according to claim 4, characterized in that, The process of controlling the scanner to scan the support structure along the refined scanning path includes: The spherical scan point set is arranged in descending order of Z value; For each of the stent centroids, the following actions are performed sequentially: Control the turntable to rotate to the motion angle corresponding to the projection of the support's center of mass onto the turntable plane; Control the three-degree-of-freedom moving mechanism to move to the partition height corresponding to the center of mass of the support; The six-degree-of-freedom robot is controlled to perform linear motion on the spherical scanning points according to the generation order of the spherical scanning points in the set of spherical scanning points.

7. The adaptive planning method for automated scanning path of cabin structure according to claim 1, characterized in that, Before the controlled scanner scans the inner surface of the cabin structure along a preset initial path, the method further includes: The cabin structure is longitudinally partitioned based on the Z-axis range of the robot's reachable space, resulting in multiple partitions; Calculate the coordinate points of the three-degree-of-freedom moving mechanism corresponding to each of the partition boundaries; Based on the theoretical shape of the cabin structure and the field of view of the scanner, the preset initial path is generated.

8. The adaptive planning method for automated scanning path of cabin structure according to claim 7, characterized in that, Before longitudinally partitioning the cabin structure within the Z-axis range of the robot-based reachable space, the method further includes: The position of the target ball is acquired using a laser tracker. A coordinate system for the turntable and a coordinate system for the three-degree-of-freedom moving mechanism are established by fitting the cylindrical surface and the plane, so as to obtain the calibration relationship between the coordinate system for the turntable and the coordinate system for the three-degree-of-freedom moving mechanism.

9. An automated scanning path adaptive planning system for cabin structures, characterized in that, include: The preliminary scan planning module is used to control the scanner to scan the inner surface of the cabin structure along a preset initial path during the preliminary scan stage. During the scan, collision detection is performed based on the depth information of the point closest to the inner surface in the current frame point cloud acquired by the scanner in real time. In response to the detection of collision risk, the pose of the scanner is dynamically adjusted radially to generate a collision-free preliminary scan path. The feature mapping module is used to register the preliminary scan point cloud obtained based on the preliminary scan path with the theoretical digital model of the cabin structure, obtain the spatial transformation relationship, and map the support centroid position in the theoretical digital model to the actual physical space based on the spatial transformation relationship. The refined scanning planning module is used to generate a spherical scanning point set with the mapped centroid of the support as the center of the sphere, and to plan a refined scanning path based on the spherical scanning point set, and control the scanner to scan the support structure along the refined scanning path.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the adaptive planning method for automated scanning path of the cabin structure as described in any one of claims 1 to 8.

Images