A method for automatic filament placement trajectory planning and design of curved surfaces with variable bandwidth and its storage medium

Abstract

This invention discloses a method for automatic fiber placement trajectory planning and design on curved surfaces with variable bandwidth, belonging to the field of automatic fiber placement in composite materials. The method includes: acquiring the surface parameters and ply parameters of the fiber placement mold; determining the reference coordinate system and starting point for each ply, generating a reference trajectory, the reference coordinate system being generated based on the admissible transformation theorem and the definition of a first-order frame field; generating a densely laid trajectory using a variable bandwidth tiling optimization algorithm based on the variable curvature parameters of the surface, and performing wide and narrow band tiling planning; and generating the final trajectory planning scheme into the system language of the placement robot for controlling the robot to perform placement work. This invention solves the problem of large gaps or overlaps in traditional free-form surface fiber placement path planning due to the limitations of the surface shape, enabling full-coverage fiber placement path planning for various curved surface structures and ensuring the accuracy of automatic fiber placement.

Description

Technical Field

[0001] This invention belongs to the field of automated fiber placement molding of composite materials, and relates to path planning in the molding process of automated fiber placement technology for composite materials. Specifically, it relates to a method for planning and designing the trajectory of variable width carbon fiber bundles with variable curvature surfaces and its storage medium. Background Technology

[0002] Automated fiber placement technology is a high-precision, high-efficiency molding and manufacturing method proposed by Boeing to meet the needs of aircraft manufacturing, and it has now been widely commercialized. Automated fiber placement technology uses fiber bundles with a narrower width than automated tape placement, thus enabling the manufacture of composite material products with more complex curved surfaces. This provides technical support for the efficient and high-precision manufacturing of high-performance composite material structures, enabling the manufacture of complex lightweight structural forms, expanding the structural forms of lightweight structures, and further broadening the application range of lightweight composite material structures.

[0003] In the automated fiber placement process, based on the surface structure of the placement mandrel, a fiber placement trajectory that completely covers the curved surface is designed for different material properties. During placement, the pre-impregnated fiber bundle is compacted and shaped using heating and compaction rollers. The entire placement process is controlled and coordinated by the robot's host computer. Automated fiber placement technology can place not only standard developable and revolution surfaces, but also complex hypercurvature and free-form surface components. Based on the boundary conditions of the layup design, the layup can be locally trimmed to achieve requirements such as thickening of the component structure, decreasing layup height, and open layup. The automated fiber placement trajectory planning technology designs algorithms for the robot's fiber placement head trajectory during processing. Based on the geometric parameters of the curved surface component, it generates a fiber placement trajectory that completely covers the entire curved surface layup, including two parts: a baseline trajectory generation algorithm and a trajectory close-layup algorithm.

[0004] Currently, the design of automatic filament placement trajectory planning technology for free-form surfaces mostly adopts algorithms based on planar approximation and trajectory point offset. The planar approximation involves a large amount of cumulative calculation on the plane during the trajectory from the starting point to the ending point, resulting in a large error in the approximate result. On irregular free-form surfaces, the equidistant offset algorithm for the trajectory points of the reference trajectory often results in a non-smooth trajectory due to curvature changes. The spline curve generated based on the trajectory points often has a large angular deviation from the initial reference trajectory, and frequent angular changes can easily cause defects such as filament deformation during the placement process. As a result, the performance of the placed product is reduced due to angular changes, and the product performance cannot meet the design requirements, making it difficult to apply to practical engineering use.

[0005] Therefore, an algorithm is needed to plan and design the filament laying trajectory for the surface of free-form mold structures, so as to achieve high-precision filament laying and processing by robots on irregular curved surfaces. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a method for planning and designing the trajectory of automatic filament placement forming on curved surfaces with variable bandwidth, along with its storage medium. Based on this method, a software program for planning the trajectory of a filament placement robot on a curved surface is developed. The method of this invention is used to generate the trajectory of a free-form surface of a current digital model, thereby enabling the robot to perform high-precision filament placement processing on irregular curved surface structures.

[0007] To achieve the above technical objectives, the present invention provides an automatic filament placement trajectory planning and design method for curved surface variable bandwidth, comprising the following steps:

[0008] S1: Establish the data model of the fiber layup mold and read the model parameters to obtain the surface parameters and layup parameters of the mold surface stored in the three-dimensional model;

[0009] S2: Determine the reference coordinate system and starting point for each ply, and generate the baseline trajectory;

[0010] S3: In each layer, a variable bandwidth tiling optimization algorithm is established based on the change of surface curvature parameters to generate tiling trajectories;

[0011] S4: Based on surface parameters and tiling algorithm, perform wide and narrow band trajectory planning with variable bandwidth tiling;

[0012] S5: The system language for generating trajectory planning schemes for laying robots, used to control the robots to perform laying work.

[0013] In step S1, the ply parameters obtained include ply boundary information, ply design sequence, and the layup angle of each ply.

[0014] In step S2, for each layup, a starting point is selected from the boundary where the layup direction of the layup intersects with the surface of the mold. A natural frame field is established at the starting point, and the natural frame field is normalized. Then, a reference coordinate system is determined based on the admissible transformation theorem. In the reference coordinate system, in addition to the normal vector, one unit direction points to the layup direction, and the other unit direction is the trajectory densification direction. The starting point is used as the starting point of the reference trajectory, and a reference trajectory is generated towards the opposite boundary. The direction of generating the reference trajectory is consistent with the layup direction.

[0015] In step S3, a variable bandwidth tiling optimization algorithm is used. Starting from the reference trajectory, the trajectory generation process is repeated until the generated trajectory reaches the ply boundary. The trajectory generation process is as follows: the curvature parameters of the surface position points are calculated based on the surface parameters; the current trajectory is divided into discrete points with unequal spacing according to the change amplitude of the curvature parameters; the offset distance of each discrete point on the surface along the trajectory densification direction is calculated, and P is the two discrete points before and after the offset corresponding to the maximum offset distance. max and P max ', the discrete point P max 'Using it as the starting point of the next trajectory, the next trajectory is generated along the ply direction.'

[0016] In step S3, the surface of the mold in the trajectory offset region is approximated as an arc with the radius of curvature of the discrete points as the radius, and the length of the arc is the width of the filament layup. The unit orthogonal frame field of the discrete points is determined, and the normal curvature of the given layup direction is calculated. There is a unique radius of curvature. The offset spacing of the discrete points on the surface along the trajectory densification direction is determined in combination with the normal radius of curvature. The point with the minimum normal curvature corresponds to the case with the maximum offset spacing.

[0017] In step S4, for two adjacent trajectories, the maximum offset distance d of the first trajectory is obtained based on the variable bandwidth tiling optimization algorithm. max As the width of the wide filament bundle, with the discrete point P of the first trajectory max Width d is determined along the normal curvature direction based on the reference. max The offset of the filament bundle; for the gap between the two tracks, select matching narrow filament bundles for close laying according to the gap allowance requirements.

[0018] Furthermore, the present invention also implements a readable storage medium storing a computer program, which, when executed by a processor, implements a method for planning and designing an automatic filament placement trajectory for curved variable bandwidth according to the present invention. The program uses G-code as the storage code; executing the program is used to control a robot to perform filament placement work.

[0019] The advantages of the method of the present invention are:

[0020] (1) The program in the method and storage medium of this invention establishes a variable bandwidth tiling optimization algorithm based on the change of curvature parameters of the surface. Through the innovation of variable width tiling, it solves the problem that the traditional free surface filament tiling path planning is limited by the surface shape and has large gaps or overlaps. It can realize full tiling path planning for various surface structures and ensure the accuracy of automatic fiber tiling.

[0021] (2) The method of the present invention is based on the mathematical method of differential geometry, which realizes the generation of reference trajectory on free surface. Moreover, based on the rotational invariance of the surface frame, it can be solved for any starting point, which effectively improves the efficiency and convenience of trajectory planning.

[0022] (3) The program in the method and storage medium of the present invention is designed based on the standard three-dimensional modeling file format, which can match and be recognized by various industrial software, effectively improving the applicability of the equipment.

[0023] (4) The method of the present invention is implemented as a set of filament laying robot curved surface variable width laying trajectory planning algorithm software, which can realize the planning and design of filament laying trajectory for dense laying on free surface, and is easy to further upgrade the algorithm and optimize the program. Attached Figure Description

[0024] Figure 1 This is a flowchart of the automatic wire placement and forming trajectory planning and design method for curved surface variable bandwidth according to the present invention.

[0025] Figure 2 This is a schematic diagram of a sample digital model of the mold used in this invention;

[0026] Figure 3 This is a schematic diagram illustrating the selection of the starting point, the establishment of the reference coordinate system, and the generation of the reference trajectory in the method of the present invention.

[0027] Figure 4 This is a schematic diagram of the initial point shift during trajectory tiling in the method of the present invention;

[0028] Figure 5 This is a schematic diagram illustrating the generation of tiling trajectories using a variable bandwidth tiling optimization algorithm in the method of this invention.

[0029] Figure 6 This is a schematic diagram illustrating the wide and narrow band trajectory planning using the variable bandwidth tiling optimization algorithm of the present invention. Detailed Implementation

[0030] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. The specific implementation methods for curved surface models described below are only used to illustrate the process of the present invention and are not intended to limit the scope of the present invention.

[0031] like Figure 1 As shown, the present invention provides an automatic wire laying trajectory planning and design method for curved surface variable bandwidth, which is described in five steps below.

[0032] S1: Read the data model of the actual mold used to lay up the composite material, and determine the parameter information of the mold surface and the layup.

[0033] Use CATIA modeling software to create a digital geometric model of the mortar placement mold, such as... Figure 2 As shown, the digital model adopts a standard CAD file format, and the surface accuracy depends on the file storage method used. In this embodiment of the invention, the three-dimensional digital model of the laying mold is saved in a format commonly used by three-dimensional modeling software, such as IGS or STP, and imported into the surface trajectory planning algorithm software for the filament laying robot developed based on the method of this invention for model processing. The module functions embedded in the developed trajectory planning algorithm software can identify and store the geometric information in the mold digital model. The identified information includes the laying surface and layup parameter data. The layup parameters include the boundary information of the layup, the design sequence of the layup, and the trajectory laying direction of each layup. The identified laying surface data includes an array storing the coordinates and normal vectors of each point on the surface.

[0034] To enable the algorithm to adapt to different fiber placement equipment and fiber bundle materials, the software provides an input port for setting fiber bundle parameters. This allows for simultaneous adaptation to various fiber bundle widths, enabling trajectory planning and design for multiple variable-width fiber bundles, and achieving trajectory planning to adapt to complex curved mold structures. The material data for the carbon fiber composite material that needs to be preset should include information such as the type of prepreg resin, fiber bundle width, and fiber bundle thickness.

[0035] S2: Determine the reference coordinate system and starting position of each ply, and generate a reference trajectory for the ply.

[0036] Based on the surface parameters of the mold obtained in step S1, the base coordinate system of the surface model, i.e. the part coordinate system, can be selected; based on the layup parameters obtained in step S1, the design boundary and extension boundary of the actual layup area of ​​each layup on the surface can be determined; based on the layup angle, the starting point position can be determined, and a reference coordinate system for the layup can be established to generate the reference trajectory for the surface layup.

[0037] Based on the existence theorem of orthogonal parametric curve networks in differential geometry:

[0038] For every point p∈S on the regular parametric surface S: r=r(u,v), there must be a neighborhood of point p. and the new parameter system on U So that in the parameter system The tangent vectors of the curves below are orthogonal to each other, that is, the parameter system... It is the orthogonal parameter system of surface S on U.

[0039] If a vector frame field with any parameter system direction can be selected at any point, and only the modulus is different, then coordinate equations can be established for any pair of orthogonal parameter systems of the curved surface mold.

[0040] The field modulus of the natural frame established for orthogonal parameter systems is arbitrary, which makes it difficult to transform with the ground coordinate system. Therefore, it needs to be normalized to facilitate the description and transformation of parameters between multiple reference systems.

[0041] Let E be the European space. 3 The parametric equation of the inner surface S is r = r(u,v), and the corresponding natural frame field is: {r; r1, r2, n}; where, n is the surface normal vector.

[0042] Let the first fundamental form of the surface S described using Gaussian notation be: I = E(du) 2 +2Fdudv+G(dv) 2 Where the Gaussian metric parameters are E = r1·r1, F = r1·r2, and G = r2·r2, then from {r1,r2}, after Schmidt orthogonalization, we obtain:

[0043] e3 = e1 × e2 = n;

[0044] The formula for surface S is as follows: Let W1 and W2 be the expressions within parentheses of the two terms in the formula. Then, for the natural frame field of the surface, we have dr = W1e1 + W2e2 and I = dr·dr. Thus, e1 and e2 in the unit orthogonal frame field {r; e1, e2, e3} on the surface S are called tangent vectors of the surface S. Such a frame field is called the first-order frame field of the surface S. At this time, e3 is linearly dependent on the normal vector n, that is, dr·e3 = 0. The rotation of the frame around e3 is arbitrary, and a first-order frame field can be established facing any wire laying direction.

[0045] This invention guarantees that any parameter system selected at any point in any direction can be determined as a unit orthogonal parameter system. That is, at any point on the surface, based on the design angle of the ply, the direction vector field of the selected parameter system can be used. Then, according to the above theorem, a first-order frame field associated with the surface and the world coordinate system is established, thereby realizing the description and expression of the ply pose information on the surface. The method of this invention determines the surface trajectory planning direction based on the ply angle data in the ply design. The default fixed angles for plying are 0°, ±45°, and 90°. Simultaneously, because the initial natural frame selection direction has multiple solutions, manifested as a rotation about the e3 axis on the first-order frame field, frame information at any angle can be obtained by solving the corresponding rotation matrix, even outside the default angle. That is, this frame system is also applicable to ply angle information of any design.

[0046] The software provides an interface for selecting and generating coordinate axes, enabling the selection of the coordinate axes. By selecting and determining the origin and the directions of the two coordinate axes, the coordinate system of the part can be constructed. It can also reuse the original model reference system based on the read data. Following the method of establishing a first-order frame field in step 2, the permissible transformation relationship between the part coordinate system and the frame field of the surface laying area is calculated. A corresponding first-order frame field for trajectory planning is established for the set lay-up direction, converting the model's pose information into trajectory pose information. Simultaneously, the model parameter system of the mold surface and the first-order frame field are comprehensively calculated to generate the initial lay-up area of ​​the surface, providing the parameter range of the lay-up area boundary in the model parameter system. By modifying the parameter range, the lay-up range can be corrected for lay-up boundaries with different lay-up variations, ensuring flexible drawing of the lay-up area based on the area size of each lay-up layer in the lay-up design, further improving the trajectory planning performance of layered variable-area lay-up.

[0047] Based on the data storage principle of a three-dimensional surface model, the information storage data group of any point on the mold surface of this invention consists of the point's coordinate information and normal vector information. This allows for the acquisition of curvature data at the selected point on the surface, expressed as average curvature and Gaussian curvature. Then, according to the layup parameters and the layup start point information, the initial path is taken as the projection curve of the tangent vector of the start point in the layup angle direction and the surface.

[0048] like Figure 3 As shown, based on the layup direction given in step S1, a starting point is selected from any position on the boundary where the layup direction of the layer intersects with the curved surface of the mold. In this embodiment of the invention, a point is arbitrarily selected within the central region of the intersecting boundary and defined as P0. P0 is used as the starting point of the reference trajectory, and a reference trajectory is generated towards the opposite boundary. The direction of generating the reference trajectory is consistent with the layup direction. Based on the parameter frame field of the mold, the natural frame field of P0, the surface normal vector n, and the parametric curves u and v, and the direction vector r can be determined. u r v A unique surface frame constituting the surface at that location can be determined, in which n and r are defined. u The intersection of the cross section and the curved surface forms the reference trajectory for the laying.

[0049] To achieve a scheme where the trajectory tessellates the entire layer, based on the reference trajectory generated in the above steps, and based on the definition of the first-order frame field given in S2, and based on the natural frame field parameters, combined with the Gaussian metric coefficients E, F, and G, the natural frame field at P0 can be normalized to obtain {r; e1, e2, e3}. Based on the admissible transformation theorem: according to the definition of the first-order frame field of a surface, the first-order frame field {r; e1, e2, e3} of surface S, rotated by any angle around e3, is still the first-order frame field of surface S. By rotating the first-order frame field about an axis, the transformation formula can be obtained:

[0050]

[0051]

[0052] Where λ is the rotation angle, the new first-order reference frame field can be expressed as: By rotating the frame by an angle λ, one unit direction of the frame points to the ply direction. At this point, one direction of the frame is the ply direction, e3 is the normal vector direction, and the third unit direction is the trajectory densification direction.

[0053] S3: In each layer, a variable bandwidth tiling optimization algorithm is established based on the change of surface curvature parameters to generate tiling trajectories.

[0054] The trajectory planning process on the surface of the mold mainly consists of two steps. The first step is to generate the reference trajectory for surface tiling, which has been explained in step S2 above. The second step is to tile the generated reference trajectory to cover the entire surface structure, which will be explained in detail below.

[0055] Based on the accuracy requirements for yarn placement trajectory recognition of the yarn placement robot, the baseline trajectory is divided into discrete points P1, P2, ..., starting from P0, until the trajectory endpoint P. n Then, using each discrete point as a reference, the position information of the corresponding discrete point on the offset trajectory is solved. The reference trajectory is divided into discrete points of the surface with unequal spacing according to the variation of the surface curvature parameters. Based on the average curvature and Gaussian curvature of each discrete point, the principal curvature of the surface structure at each discrete point can be determined, and the orthogonal frame direction based on the principal curvature direction can be determined. Further, the local surface differential processing can be used to determine the filament laying width. Within a local range, the surface can be fitted as an arc with the curvature radius as the radius, and the arc length is the filament laying width. For all discrete points belonging to the same trajectory, the normal curvature is compared, and the point with the minimum normal curvature is the reference for generating the next trajectory. In this invention, the magnitude of the discrete point offset distance is solved based on the formula of differential surface arc and offset, and the curvature radius is determined as the main influencing factor of the offset. Based on the normal curvature determined in step S3, there is a unique corresponding curvature radius, thereby obtaining the discrete point offset spacing.

[0056] like Figure 4 As shown, the initial point P0 is offset on the curved surface along the trajectory densification direction. In one offset, since the offset distance is much smaller than the mold size, the mold surface of the trajectory offset region can be approximated as an arc with the radius of curvature of the initial offset point as the radius. Then, according to the principle of vector calculation, the offset spacing d can be obtained as follows:

[0057] d = Rsin(θ+β) - Rsinβ;

[0058] θ = ω / R;

[0059] According to Euler's formula, the normal curvature k is obtained as follows:

[0060]

[0061]

[0062] Where θ is the angle of rotation from the offset distance d at P0 to the next point, ω is the offset distance during one offset, and ω is a preset offset distance value; R is the radius of curvature of the starting offset point, such as... Figure 4 As shown, β is the angle between the projection line of P0 onto the part coordinate system plane and the radius of curvature; k1 and k2 are the principal curvatures in the two directions e1 and e2, respectively. Let v1 be the angle between the normal curvature k and the principal curvature k1, and v1 be the direction vector of the principal curvature corresponding to k1. Therefore, the normal curvature radius R at that point can be calculated.

[0063]

[0064] The value of β is determined by the coordinates of P0 and the center coordinates of the equivalent circular arc. In the model coordinate system, both points are uniquely determined. Therefore, for a P0 point with a defined coordinate, there exists a defined value of β, which means that the offset spacing d can be obtained.

[0065] like Figure 5 As shown, a reference point is obtained on the second trajectory, including: for P0, P1, P2...P n The discrete points are calculated using the same process as above for calculating the offset spacing d, to obtain the frame field of each point and the reference points P0', P1', P2'…P of the second trajectory. n ', compare each discrete point P i and P i The offset intervals d0, d1, d2…d between '(i=0,1,2,…n)' are: n Let the maximum offset spacing d be among them. max The discrete points of the corresponding two trajectories are P and P respectively. max and P max ', the maximum offset spacing d max The corresponding discrete point P max 'This serves as the starting point for the second trajectory. Then, the second trajectory is generated along the layup direction, and discrete points are selected using the same steps as the baseline trajectory. The trajectory offset algorithm is repeated at these discrete points to determine the third trajectory. This process is repeated until the mold boundary conditions are met. The filament layup width between two adjacent trajectories is the calculated maximum offset distance. 'n' is a positive integer used to mark the number of segments into which the trajectory is divided.

[0066] S4: Based on surface parameters and variable bandwidth tiling optimization algorithm, perform wide and narrow band trajectory planning for variable bandwidth tiling.

[0067] Based on the offset spacing calculation formula in step S3, for all discrete points of a single trajectory, determine the discrete point with the smallest offset distance. Using this point as a reference, offset the width of the wide filament bundle along the normal curvature direction. For the discrete point with the largest offset distance, calculate the offset distance difference between it and the discrete point with the smallest offset distance. According to the gap allowance requirements, select matching narrow filament bundles to densely fill the entire layup.

[0068] The width of the wide filament bundle is offset based on the closely spaced trajectory generated in step S3. For two adjacent trajectories, the maximum offset spacing d is calculated from the discrete points of the first trajectory. max The width of the wide filament bundle is given, and the generated trajectory is the centerline of the wide filament bundle, with d on both sides of the centerline. max / 2 is the boundary of this filament bundle, that is, the gap boundary; it is necessary to densely lay narrow filament bundles in the gap between two adjacent tracks.

[0069] like Figure 6 As shown, for two adjacent trajectories, the offset distance d calculated based on the above steps... i For i = 0, 1, 2, ..., n, compare the distance d between each discrete point. i and the maximum spacing d used max The difference △d i The distribution of △d i =d max -d i The difference Δd is obtained. i The distribution pattern, assuming at discrete point P i The calculated offset spacing is d i The corresponding discrete point after the offset is P. i ', at spacing d i The midpoint position, i.e., the discrete point P. i With P i The midpoint of ' is defined as the discrete point Q of the tiling. i (i = 0, 1, 2, 3…n), move towards the midpoint Q i The distance difference Δd between the two sides is obtained. i / 2 has two endpoints, and the variable-width gap Δs between the two endpoints is a discrete point. i (i = 0, 1, 2, 3…n). Lay out discrete points Q. i The narrow filamentary trajectory is formed by i = 0, 1, 2, 3…n.

[0070] Based on △s iBased on the data distribution and layup design scheme, the narrow filament width range was modified to ω′, ensuring that the coverage and minimum gap meet the design requirements after laying filaments with a width of ω′. ω′ takes a value greater than 0 and not greater than Δd. i The maximum value 'a' in the data, and the gap allowance requirements are met. When performing close-laying of narrow filament bundles, the screening gap difference Δs... i The position point Q where the filament bundle with width ω′ is laid out. j (j∈{0,1,2,3…n}), the selection criterion is to satisfy the condition that no stacking occurs, i.e., Δs i The location point within the range [ω′, a]. Determined by the discrete point Q. j Narrow filament bundles are densely laid on both sides within the two endpoints along the direction of normal curvature.

[0071] The above yielded the wide and narrow filament bundle dense tiling trajectory planning, thus enabling a variable width dense tiling method for the entire curved surface mold.

[0072] S5: Generates the robot control language from the laying trajectory scheme, which is used to control the robot to perform the laying work.

[0073] The trajectory planning and design method generated by this invention is implemented as a robot-recognizable G-code file. Based on the trajectory planning scheme obtained from the above steps, the sequence and geometric parameter information of all trajectories on the curved surface are obtained. According to the robot programming language, the parameter information is converted into robot control language such as G-code to control the robot to perform the filament laying operation.

[0074] Furthermore, the present invention provides a readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for planning and designing an automatic filament placement trajectory for curved surfaces with variable bandwidth. The program uses G-code as storage code and generates one or more stored document files based on the upper limit of the file size that the robot can read.

[0075] In summary, this invention proposes a curved surface variable bandwidth automatic fiber placement trajectory planning and design method and its storage medium for carbon fiber automated fiber placement robot trajectory planning. This method overcomes the limitations of traditional trajectory planning methods, which are restricted by curved surface structures, leading to problems such as inability to achieve accurate trajectory placement or deformation caused by incomplete overlap between the placement process and the curved surface structure, resulting in defects. It also ensures the accuracy of automatic fiber placement. Furthermore, this invention provides software for trajectory planning based on a standard 3D digital model program, which can be read and recognized by various software programs, meeting the trajectory planning design requirements for various fiber bundle specifications and different placement head specifications. The curved surface trajectory planning algorithm software program for the fiber placement robot in the storage medium of this invention can generate path data for curved surface trajectory planning and further convert it into a mechanical language output scheme. It can also perform robot curved surface fiber placement trajectory planning design under different fiber bundle widths and thicknesses and different placement head specifications, providing technical guidance for practical robot placement optimization schemes.

[0076] Except for the technical features described in the specification, all other technologies are known to those skilled in the art. Descriptions of well-known components and technologies are omitted in this invention to avoid redundancy and unnecessary limitation. The embodiments described above do not represent all embodiments consistent with this application. Various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this invention are still within the protection scope of this invention.

Claims

1. A method for planning and designing the trajectory of automatic filament placement for curved surfaces with variable bandwidth, characterized in that, Includes the following steps: S1: Establish a digital model of the layup mold, obtain the surface parameters of the mold surface, and the layup parameters; the obtained layup parameters include the design sequence of the layup, the boundary information of the layup, and the layup angle of each layup layer; S2: Determine the reference coordinate system and starting point for each ply, and generate the baseline trajectory; For each layup, select the starting point from the boundary where the layup direction intersects with the curved surface of the mold. A natural frame field is established at the starting point. The natural frame field is normalized, and a reference coordinate system is determined based on the admissible transformation theorem. In the reference coordinate system, except for the normal vector, one unit direction points to the ply direction, and the other unit direction is the trajectory densification direction. The starting point is used as the starting point of the reference trajectory, and the reference trajectory is generated towards the opposite boundary. The direction of generating the reference trajectory is consistent with the ply direction. Let P0 be the starting point for the current layup. Based on the parametric frame field r(u,v) of the mold surface S, the natural frame field {r;r1,r2,n} at P0 is determined, where... , , , where n is the surface normal vector; by combining the Gaussian metric coefficients E, F, and G, the natural frame field at P0 is normalized to obtain {r; , , }; where E= F= G= , = , , = =n;Unit orthogonal frame field{r; , , In} , It is the tangent vector of the surface S, and the unit orthogonal frame field is called the first-order frame field of the surface S; Based on the admissible transformation theorem: the first-order frame field {r;} of surface S: , , } around Rotating it by any angle still results in a first-order frame field on the surface S; therefore, the first-order frame field is rotated around... Rotation angle A new first-order frame field {r; is obtained. , , },at this time, and One unit direction points to the ply direction, and the other unit direction is the trajectory densification direction; the new first-order frame field obtained is the reference coordinate system established for the current ply; S3: In each layer, a variable bandwidth tiling optimization algorithm based on the change of surface curvature parameters is used to generate the tiling trajectory. In the aforementioned variable bandwidth tiling optimization algorithm, starting from the reference trajectory, the trajectory generation process is repeated until the generated trajectory reaches the ply boundary. The trajectory generation process is as follows: the curvature parameters of the surface position points are calculated based on the surface parameters; the current trajectory is divided into discrete points with unequal spacing according to the change amplitude of the curvature parameters; the offset distance of each discrete point on the surface along the trajectory densification direction is calculated, and the maximum offset distance d is set. max The corresponding discrete points before and after the offset are P max and P max ', the discrete point P max 'Using it as the starting point of the next trajectory, the next trajectory is generated along the ply direction;' The method for calculating the offset spacing of discrete points on the curved surface along the trajectory densification direction is as follows: Assume a discrete point P0 is offset on the curved surface along the trajectory densification direction. Approximate the mold surface of the trajectory offset region as an arc with the radius of curvature of discrete point P0 as its radius, and the length of the arc is the width of the filament bundle. Calculate the curvature parameters based on the coordinates and normal vector of P0, and determine the unit orthogonal frame field {r;} at P0. , , } and principal curvature, assuming corresponding , The principal curvatures are k1 and k2, and the corresponding normal vectors are... The normal curvature is k, as follows: , Where φ is the angle between the normal curvature k and the principal curvature k1, and v1 is the principal curvature direction vector corresponding to k1; Then calculate the normal radius of curvature at P0. Further obtain the offset distance d of discrete point P0 on the surface along the trajectory densification direction, which is d = Rsin(θ+β) - Rsinβ, where θ = ω / R; where θ is the angle rotated from the offset distance d at P0 to the next point, ω is the preset offset distance value, and β is the angle between the projection line of P0 onto the part coordinate system plane and the radius of curvature. S4: Wide and narrow band trajectory planning based on variable bandwidth tiling optimization algorithm; For two adjacent trajectories, the maximum offset spacing d of the first trajectory will be obtained based on the variable bandwidth tiling optimization algorithm. max As the width of the wide filament bundle, with the discrete point P of the first trajectory max Width d is determined along the normal curvature direction based on the reference. max The offset of the filament bundle; for the gap between the two tracks, select matching narrow filament bundles for close laying according to the gap allowance requirements; After two adjacent tracks are laid using wide filament bundles, the gap between the two tracks is then filled with narrow filament bundles as follows: Let P be the discrete point of the first track among the two adjacent tracks. i The calculated offset spacing is d i Calculate d i and the maximum offset spacing d max The difference △d i Given i = 0, 1, 2, ..., n, and n+1 discrete points; obtain the discrete point P. i Offset spacing d i The position of / 2 is defined as the location of the discrete point Q. i , i=0,1,2,…n; by laying out discrete points Q i Obtain the distance difference Δd between the two sides along the normal curvature direction. i The two endpoints of / 2 require a narrow filament bundle to be densely laid between the two endpoints, i=0,1,2,…n; the width of the narrow filament bundle is selected. , The value is greater than 0 and not greater than Δd. i The maximum value in the range, and meets the gap allowance requirements; from the layup discrete point Q i Filter out the difference △d i Not less than discrete point Q j j∈{0,1,2,3…n}, by laying out discrete points Q j Narrow filament bundles are densely laid on both sides within the two endpoints along the direction of normal curvature.

2. The method according to claim 1, characterized in that, In step S1, the tow material data is also preset, including the type of prepreg resin, tow width, and tow thickness.

3. The method according to claim 1, characterized in that, In step S2, the starting point is selected from the center region of the intersecting boundaries.

4. The method according to claim 1, characterized in that, The method described above generates the system language for the laying trajectory planning scheme, which is used to control the robot to perform the filament laying work.

5. A readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the automatic filament placement trajectory planning and design method for curved surface variable bandwidth as described in any one of claims 1-3, wherein the program uses G-code as storage code.

Images