Array type hanging spraying robot layout optimization design method based on spraying track constraint

Abstract

The application discloses a layout optimization design method of array type hanging spraying robot based on spraying track constraint, belongs to the field of spraying technology, and comprises the following steps: discretizing a three-dimensional model of an airplane to be sprayed to obtain a point cloud feature vector set; array type spraying robot units are arranged after clustering division to form an array; a spraying torch is simulated to form a circular spraying area covering the area to be sprayed on the surface to be sprayed; the divided area is traversed according to the full coverage principle; the inverse solution of the angle of each axis of the mechanical arm is obtained by using the inverse method of the Jacobian matrix, and if there is no solution, the clustering is re-divided; finally, a fitness function is calculated, and whether the set size requirement is met is judged; if the set size requirement is not met, the clustering parameters are optimized by using a particle swarm optimization algorithm until the maximum iteration number is reached, and the optimal layout and spraying track are output. According to the application, the layout of the array type hanging spraying robot is optimized according to the structural features of the airplane, the spraying efficiency and the equipment flexibility are simultaneously improved on the basis of ensuring spraying accessibility, the application is suitable for different airplane models and can be reused.

Description

Technical Field

[0001] This invention relates to the field of spraying robot equipment technology, and more specifically to a layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints. Background Technology

[0002] Spray painting robots offer advantages such as high efficiency, stable quality, and low maintenance costs. They effectively solve the problem of manual spray painting, where subjective operational errors lead to inaccurate control of spraying angle and speed, thus hindering the assurance of coating thickness and quality. In existing technologies:

[0003] Chinese patent CN114950805B provides a method for optimizing the positioning of a robotic painting system for aircraft wings. This method calculates the robot painting area on the extended surface of the aircraft wing at different positioning points, and obtains the intersection area by finding the intersection of the robot painting area and the actual painting area of ​​the aircraft wing. This achieves full coverage painting of the aircraft wing with the fewest possible positioning points, thereby reducing seams in the painting area and ensuring the final painting quality. Although this invention improves upon the positioning optimization of the painting robot, it still has the following shortcomings:

[0004] 1) Existing technologies have proposed optimization methods for robot spraying positions, but lack index-based evaluation of these methods, meaning that the results cannot be quantitatively evaluated; 2) Considering only the sequential spraying method of a single robotic arm results in low work efficiency, making it difficult to meet the requirements of rapid spraying. Summary of the Invention

[0005] To address the aforementioned problems in existing spraying robot layout methods, this invention proposes an array-type suspended spraying robot layout optimization design method based on spraying trajectory constraints. While ensuring spraying accessibility, the layout of the array-type suspended spraying robot is optimized according to the structural characteristics of the aircraft, thereby achieving a simultaneous improvement in spraying efficiency and equipment flexibility.

[0006] To achieve the above-mentioned objectives, the technical solution of the present invention is as follows:

[0007] A layout optimization design method for array-type overhead spraying robots based on spraying trajectory constraints includes:

[0008] a. Import the 3D model of the aircraft to be painted and define the area to be painted. Discretize the area to be painted and obtain the set of all discrete point cloud feature vectors.

[0009] b. Initialize the number of cluster centers, and divide the discrete point cloud feature vector set using clustering methods to obtain the point cloud division region;

[0010] c. Arrange the motion modules of the spraying robot units according to the point cloud division results to form a spraying robot array;

[0011] d. The simulated spraying robot array starts the spray gun to form a spray torch and forms a circular spraying area on the surface to be sprayed. During spraying, the area is divided by traversing each point cloud according to the principle of full coverage.

[0012] e. For each circular spraying area on the surface to be sprayed within the divided area, traverse its corresponding module position and inversely solve its six-axis robotic arm coordinates for each axis angle at each corresponding circular spraying area position; if a corresponding solution exists, store the path of the divided area; otherwise, re-initialize the clustering parameters randomly to divide the feature set.

[0013] f. For the partitioned region path with a corresponding solution, calculate the fitness function and determine whether it meets the set size requirement. If it does, end the optimization process; otherwise, use the optimization algorithm to optimize the clustering parameters until the set maximum number of iterations is reached, and output the optimal structural layout and corresponding spraying trajectory.

[0014] Furthermore, obtain the set of feature vectors for all discrete point clouds, including: establishing a coordinate system at the aircraft's center of gravity. , O This represents the origin of the coordinate system that coincides with the centroid. Along the longitudinal axis of the aircraft Along the transverse axis of the aircraft, Along the vertical axis of the aircraft; then any discrete point cloud in the area to be painted is formed by... Perform feature representation, where Represents spatial coordinates, The normal vector is represented by an angle; ultimately, all discrete point cloud feature vectors form a set. .

[0015] Furthermore, the motion modules of each painting robot unit are arranged according to the inter-class boundaries based on the point cloud partitioning results. The inter-class boundaries are the set of boundary points, and boundary point B is defined as:

[0016] ;

[0017] in, Represents the spatial coordinates of the boundary points; Represents the boundary of the spraying area. This indicates an indicator function that, when its built-in condition is met... =1, otherwise =0; Point The neighborhood point set, Point Clustering categories.

[0018] Furthermore, the objective function of the clustering method is defined as follows:

[0019] ;

[0020] in, c For the set of cluster centers, k Let J(c) be the number of cluster centers, and J(c) be the distance from the feature vector of the discrete point cloud to the feature vector of the cluster center. p For discrete point cloud feature vectors, c i For the first i Each cluster center feature vector.

[0021] Furthermore, the solution method for the angles of each axis of the six-axis robotic arm at each corresponding circular spraying area position is as follows: traverse the module positions corresponding to each divided area, substitute the six-axis robot parameters including link length, torsion angle, offset and end target position into the Jacobian matrix inversion method, and solve the angles of each axis of the six-axis robotic arm at each circular spraying area position inversely.

[0022] Furthermore, the fitness function is calculated according to the following formula:

[0023] ;

[0024] in, For the fitness function, This indicates the degree of change in the angle of each axis when the robotic arm performs a spraying operation along a prescribed trajectory. This represents the set of angles for each axis of the robotic arm across all defined regions. This indicates the calculation of the information entropy of the set. The scaling factor is used to control the contribution between the two factors, and k is the number of cluster centers.

[0025] Furthermore, the spraying robot unit includes a motion module, a six-axis robot, an atomizing spray gun, and a paint tank; the six-axis robot and the paint tank are mounted on the motion module to achieve movement in the XY plane; the atomizing spray gun is mounted on the front flange of the six-axis robot, and the paint tank is used to provide paint during the spraying process.

[0026] Furthermore, the spraying parameters that need to be controlled during simulated spraying include: spraying flow rate q, spray gun speed v, spraying distance h, film thickness T(x), and spray width w.

[0027] Furthermore, the optimization algorithm is a particle swarm optimization algorithm.

[0028] In summary, the present invention has the following advantages:

[0029] 1. This invention optimizes the layout of the array-type hanging spraying robot based on the structural characteristics of the aircraft and the spraying trajectory, while ensuring the accessibility of the spraying process. It can be adapted to the structural characteristics of different aircraft models, achieving simultaneous improvement in spraying efficiency and equipment flexibility. It can also be reused for different aircraft models and has the advantage of low cost.

[0030] 2. This invention verifies the reachability of the robotic arm by discretizing the area to be sprayed, clustering the point cloud, and using the Jacobian matrix inversion method. Combined with the principle of full coverage spraying, the trajectory is planned, which can effectively ensure the rationality of the spraying trajectory and the full coverage of the spraying area, reduce spraying defects, and ensure the stability of coating thickness and spraying quality.

[0031] 3. This invention constructs a dedicated fitness function as the basis for optimization judgment, and uses the number of divided regions and the degree of change of the angle of each axis of the robotic arm as optimization indicators. It also uses the particle swarm optimization algorithm to optimize the clustering hyperparameters, so that the robot array layout design is more in line with the requirements of the spraying process. Attached Figure Description

[0032] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments, wherein:

[0033] Figure 1 This is a flowchart illustrating the optimized design of the present invention;

[0034] Figure 2 This is a schematic diagram of the point cloud division results on the upper surface of the aircraft according to the present invention;

[0035] Figure 3 This is a schematic diagram of the smallest spraying robot unit in this invention;

[0036] Figure 4 This is a schematic diagram of the boundary distribution between classes of the spraying robot array motion module of the present invention;

[0037] Figure 5 This is a schematic diagram of the spraying process;

[0038] Figure 6 This is a schematic diagram of the full-coverage spraying path. Detailed Implementation

[0039] To more clearly illustrate the present invention, the following description, in conjunction with preferred embodiments and accompanying drawings, further clarifies the invention. Those skilled in the art should understand that the specific description below is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the present invention.

[0040] For reference Figure 1 The flowchart shown illustrates the optimization design process. This invention provides a layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints, comprising the following steps:

[0041] Step 1: Import the 3D model of the aircraft to be painted.

[0042] Step 2: Define the area to be sprayed; in this implementation, the area to be sprayed is defined as the upper surface area of ​​the aircraft.

[0043] Step 3: Discretize the area to be painted in the 3D model of the aircraft to obtain a set of discrete point cloud feature vectors.

[0044] In practice, a coordinate system is established at the aircraft's center of gravity. ,in, O This represents the origin of the coordinate system that coincides with the centroid. The pitching motion of an aircraft is along its longitudinal axis, that is, the nose of the aircraft pitching up or down. The aircraft rolls around its lateral axis, meaning the fuselage tilts to the left or right. Assuming the aircraft yaws around its vertical axis (i.e., the nose turns left or right), then any discrete point cloud... m can be Perform feature representation, where Represents spatial coordinates, The normal vector is represented by an angle; ultimately, all discrete point cloud feature vectors form a set. .

[0045] Step 4: Preset the number of cluster centers on the upper surface of the aircraft, and divide the discrete point cloud feature vector set according to the clustering method to obtain the point cloud division result of the upper surface of the aircraft.

[0046] The objective function of a clustering method can be defined as:

[0047] ;

[0048] in, c Let J(c) be the set of feature vectors of cluster centers, k be the number of cluster centers (initial value of k can be set to 2 or 3), and J(c) be the distance from the feature vectors of the discrete point cloud to the feature vectors of the cluster centers. p For discrete point cloud feature vectors, c i For the first i Each cluster center feature vector.

[0049] like Figure 2 As shown, based on the preset number of cluster centers, five partitioned regions corresponding to the point cloud partitioning results on the upper surface of the aircraft can be obtained.

[0050] Step 5: Arrange the smallest spraying robot units according to the point cloud division results to form a spraying robot array.

[0051] like Figure 3The diagram shows the smallest spraying robot unit, which includes a motion module, a six-axis robot, an atomizing spray gun, and a paint tank. The six-axis robot and paint tank are mounted on the motion module, enabling movement in the XY plane. The atomizing spray gun is mounted on the front flange of the six-axis robot, and the paint tank supplies paint during the spraying process.

[0052] The motion modules of each painting robot unit are arranged according to the inter-class boundaries based on the point cloud partitioning results. The inter-class boundaries are the sets of boundary points. It can be defined as:

[0053] ;

[0054] in, Indicates the spatial coordinates of boundary point B; This represents the boundary of the spraying area defined in step two. This indicates an indicator function that, when its built-in condition is met... =1, otherwise =0; Point The neighborhood point set, Point The clustering category. That is, when the point When the cluster of points at the boundary of a spraying area or its neighboring points has a class that is not equal to its own, that point is defined as a boundary point. Multiple boundary points form inter-class boundaries, such as... Figure 4 The diagram shows the boundary distribution between the motion modules of the painting robot array.

[0055] Step Six: Simulate the activation of the spray gun to form a spray torch and create a circular spray area on the surface to be sprayed. During spraying, follow the principle of full coverage of the area to be sprayed, traversing each designated area, such as... Figure 6 As shown. Figure 5 As shown, the spraying parameters that need to be controlled include: spraying flow rate q, spray gun speed v, spraying distance h, film thickness T(x), and spray width w. The parameter settings are determined by experimental results.

[0056] Step 7: For each circular spraying area on the surface to be sprayed within the defined area, traverse the module positions corresponding to that area. Obtain the robot parameters based on the selected six-axis robot, including link length, torsion angle, offset, and end-effector target position. Substitute the six-axis robot parameters into the Jacobian matrix inversion method to inversely solve the coordinates of the six-axis robot arm at the _____. a The angles of each axis at the location of the circular spraying area If there is a corresponding analytical solution for the angle of each axis of the robotic arm coordinate system corresponding to the position of each circular spraying area, then the path of the divided area is stored; if there is no corresponding solution, then the clustering parameters are randomly initialized again to divide the feature set.

[0057] Step 8: Determine if the region partitioning and traversal are complete. If so, calculate the fitness function using the following formula:

[0058] ;

[0059] in, For the fitness function, This indicates the degree of change in the angle of each axis when the robotic arm performs a spraying operation along a prescribed trajectory. This represents the set of angles for each axis of the robotic arm across all defined regions. This indicates the calculation of the information entropy of the set. The scaling factor is used to control the contribution between the two factors.

[0060] The smaller the fitness function, that is, the fewer the number of regions and the smaller the change in the angle of each axis, the better the design tends to be.

[0061] Step 9: If the fitness function does not meet the set size requirement, use the particle swarm optimization algorithm to adjust the number of cluster centers. k The optimization process is performed, and it is determined whether the maximum number of iterations has been reached. If it has, the optimization design process ends, and the optimal structural layout and corresponding spraying trajectory are output.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications or equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints, characterized in that, include: a. Import the 3D model of the aircraft to be painted and define the area to be painted. Discretize the area to be painted and obtain the set of all discrete point cloud feature vectors. b. Initialize the number of cluster centers, and divide the discrete point cloud feature vector set using a clustering method to obtain the point cloud region division; the objective function of the clustering method is defined as follows: ； in, c For the set of cluster centers, k Let J(c) be the number of cluster centers, and J(c) be the distance from the feature vector of the discrete point cloud to the feature vector of the cluster center. p For discrete point cloud feature vectors, c i For the first i Cluster center feature vectors; c. Deploy the motion modules of the painting robots according to the point cloud region division to form a painting robot array; the motion modules of each painting robot are arranged according to the inter-class boundaries of the point cloud region division, and the inter-class boundaries are the set of boundary points, where boundary point B is defined as: ； in, Represents the spatial coordinates of the boundary points; Represents the boundary of the area to be painted. This indicates an indicator function that, when its built-in condition is met... =1, otherwise =0; Point The neighborhood point set, Point Clustering categories; d. The simulated spraying robot array starts the spray gun to form a spray torch and forms a circular spraying area on the surface of the area to be sprayed. During spraying, the area is divided by traversing each point cloud according to the principle of full coverage. e. For each circular spraying area on the surface of the area to be sprayed within the point cloud division area, traverse the position of its corresponding motion module and inversely solve the six-axis robot arm coordinates at the angles of each axis in each circular spraying area; if there is a corresponding solution, store the division area path; otherwise, re-initialize the clustering parameters randomly and divide the area. f. For the partitioned region path with a corresponding solution, calculate the fitness function and determine whether it meets the set size requirement. If it does, end the optimization process; otherwise, use the optimization algorithm to optimize the clustering parameters until the set maximum number of iterations is reached, and output the optimal structural layout and corresponding spraying trajectory.

2. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 1, characterized in that, Obtain the set of feature vectors for all discrete point clouds, including: establishing a coordinate system at the aircraft's center of gravity. , O This represents the origin of the coordinate system that coincides with the centroid. Along the longitudinal axis of the aircraft Along the transverse axis of the aircraft, Along the vertical axis of the aircraft; then any discrete point cloud in the area to be painted is formed by... Perform feature representation, where Represents spatial coordinates, The normal vector is represented by an angle; ultimately, a set of all discrete point cloud feature vectors is formed. .

3. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 1, characterized in that, The method for solving the coordinates of the six-axis robotic arm in each corresponding circular spraying area is as follows: traverse the positions of the motion modules corresponding to each divided area, substitute the six-axis robot parameters, including link length, torsion angle, offset, and end-target position, into the Jacobian matrix inversion method, and solve the coordinates of the six-axis robotic arm in each circular spraying area inversely.

4. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 1, characterized in that, The fitness function is calculated using the following formula: ； in, For the fitness function, This indicates the degree of angle change of each axis when a six-axis robotic arm performs a spraying operation according to a specified trajectory. This represents the set of angles for each axis of a six-axis robotic arm across all defined regions. This represents the information entropy of the set of angles for each axis of a six-axis robotic arm. The scaling factor is used to control the contribution between the two factors, and k is the number of cluster centers.

5. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 1, characterized in that, The painting robot includes a motion module, a six-axis robot, an atomizing spray gun, and a paint tank. The six-axis robot and the paint tank are mounted on the motion module to move in the XY plane. The atomizing spray gun is mounted on the front flange of the six-axis robot, and the paint tank is used to supply paint during the painting process.

6. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 5, characterized in that, The spraying parameters that need to be controlled during the spraying process include: spraying flow rate q, spray gun speed v, spraying distance h, film thickness T(x), and spray width w.

7. The layout optimization design method for an array-type suspended spraying robot based on spraying trajectory constraints as described in claim 1, characterized in that, The optimization algorithm is the particle swarm optimization algorithm.

Images