Robot type automatic laying path generation and post-processing simulation system and simulation method

Abstract

The invention discloses a path generation and post-processing simulation system and method for automatic laying of a robot type composite material, and aims to efficiently plan a laying path according to component design requirements, realize post-processing and simulation of the path by using a RoboDK platform, and realize component forming of the composite material. According to the system, an STL file is imported through a model definition module, and the geometric dimension and the structure of a to-be-laid curved surface are defined; the initial path generation module determines a reference line, a path starting point and a laying layer angle of a laying path, determines a laying direction through a distance calculation and projection method, and calculates an intersection point with a triangular patch boundary; generating a series of path points through multiple iterations, and performing path fitting by using a cubic B spline; densifying the initial path by using a parallel equidistant method to ensure that the curved surface to be laid is covered in the laying process; the generated path is output in a CSV file form, simulation verification is carried out through RoboDK, and it is ensured that the posture of the laying head is consistent with a workpiece coordinate system.

Description

Technical Field

[0001] The present invention belongs to the field of composite material forming processes, and particularly relates to a robot - type automatic placement path generation and post - processing simulation system and simulation method for curved surface components. Background Art

[0002] Fiber - reinforced composite materials are high - performance materials composed of reinforcing fibers and resin matrices. Due to their excellent specific strength, specific modulus, fatigue resistance, and corrosion resistance, as well as good designability, they are widely used in fields such as aerospace, new energy vehicles, and wind power generation. However, traditional manufacturing methods for composite components usually rely on techniques such as manual placement, winding, and sewing. These methods have some significant limitations, including a small working range, poor positioning accuracy, low placement efficiency, insufficient precision in controlling placement pressure, and inability to accurately guarantee placement thickness and angle, resulting in unstable quality and high labor intensity.

[0003] In contrast, the automatic fiber placement technology, as an emerging method for composite material automation processing, has advantages such as low cost, high efficiency, and high material utilization rate, and can achieve flexible automation in production and manufacturing. The difference between this technology and traditional machining is that it adopts an additive manufacturing mode of layer - by - layer stacking and is placed according to the designed direction.

[0004] The automatic placement path generation and post - processing prevention system generates corresponding robotic arm (RAPID) processing codes based on the material placement process characteristics and component shape features, combined with the structural design requirements of the component and the machine structure and working mode of the placement equipment, to guide the dedicated placement equipment to realize the forming and manufacturing of composite components. On the premise of meeting the design requirements, this system aims to improve placement efficiency, ensure placement quality, save materials, and reduce manufacturing costs.

[0005] Domestic universities and other institutions have carried out a series of research and experiments in various modules of placement software (such as trajectory planning technology) and made certain progress. However, the current domestic research mainly focuses on the development based on CATIA for certain functional modules, and there is still a large gap from forming a stable, reliable, and efficient placement software. In addition, during the path densification process, most of them are the beam bandwidth before the parallel equidistant beam placement at present. However, the beam bandwidth will increase to a certain extent after placement, resulting in an increase in the defects of gaps and overlaps during the placement process. Therefore, it is of great significance to use the bandwidth size after placement to feedback the offset distance during the placement process to reduce the defect rate of gaps and overlaps of the placed components. Summary of the Invention

[0006] The object of the present invention is to overcome the deficiencies of the prior art and provide a robotic automatic placement path generation and post-processing simulation system and simulation method on a curved surface, which can realize intelligent automatic fiber placement, improve the operation efficiency, placement quality, save materials, and reduce the manufacturing cost of a robotic automatic fiber placement machine.

[0007] The present invention provides the following technical solutions:

[0008] A robotic automatic placement path generation and post-processing simulation system includes a model definition module, an initial path generation module, a path densification module, and a path post-processing simulation module;

[0009] The model definition module is used to define the geometric model data of the surface to be placed;

[0010] The initial path generation module is used to define the parameters of the placement path, calculate and generate path points in combination with the defined parameters, and fit the generated path points into the required initial path;

[0011] The path densification module uses the parallel equidistant method to densify the initial path;

[0012] The path post-processing simulation module outputs the planned curves in the densified path, an automatic placement simulation platform, and performs automatic placement simulation. After the simulation is verified to be correct, automatic placement is carried out.

[0013] Further, the initial path generation module combines the defined model parameters, calculates the triangular patch unit where the path start point is located, calculates the projection of the reference direction under the corresponding triangular patch, and calculates the intersection point of the path start point along the projection direction and the boundary of the corresponding triangular patch unit; defines the calculated intersection point as the new start point; repeats the intersection point calculation process until the generated path points exceed the defined boundary conditions; fits a set of generated path points into the required initial path through cubic B-spline.

[0014] Further, the path densification module uses the parallel equidistant method to densify the path. The bandwidth of the fiber bundle after placement under specific working conditions is measured through experiments. Based on the initial path, the geodesic direction is solved along the tangential direction of the initial path point in the placement direction. The point that is at a distance equal to the bandwidth length obtained from the experiment along the geodesic direction and from the initial path is solved. The path points of all parallel equidistant paths are fitted, and this curve is used as the new initial path to repeat the steps until the surface boundary.

[0015] Further, the path post - processing simulation module outputs the planned curve through a csv file, builds a composite material automatic placement simulation platform based on RoboDK, imports the generated csv data point file according to the established tool and workpiece coordinate systems, conducts automatic placement simulation, and outputs the Rapid program for automatic placement after verification.

[0016] A simulation method for a robot - type automatic placement path generation and post - processing simulation system includes the following steps:

[0017] Step 1: The model definition module defines the geometric dimensions and structure of the surface to be placed by importing the model, and the model file is an STL file; by reading the information of the STL file, the information on each triangular patch element is obtained.

[0018] Step 2: The initial path generation module defines the reference line, path starting point, and ply angle of the placement path, traverses the triangular patch elements to retrieve the triangular patch element where the current starting point P is located; calculates the projection direction d′ of the reference line direction d under the corresponding triangular patch, and based on the projection direction d′, calculates the actual placement direction v, where the angle between v and d′ is the set ply angle; passes through point P along the direction v to solve its intersection point P with the boundary of the triangular patch. 1 ;

[0019] Step 3: P 1 is the next placement point of the initial path, defines P 1 as the path starting point of the next patch element, and repeats Step 2 until the placement boundary to obtain a set of path points {P 0 , P 1 , P 2 , P 3 ……}, and obtains the initial path through cubic B - spline fitting.

[0020] Step 4: When the process parameters of automatic placement are known, test the bandwidth w of the prepreg with a specified bandwidth after placement and forming, and densify the initial path using the parallel equidistant method; discretize the initial path to obtain the binormal vector e of the path at each discrete point. 2 , take the tested bandwidth w as the offset distance, and generate a geodesic segment with a specified starting point, direction, and length according to the offset distance w and the binormal vector direction e. 2 When the length of the geodesic segment is less than the set threshold L, abandon this position and repeat the step of generating the geodesic segment at the next discrete point; when the length of the geodesic segment is equal to the set threshold L, the end point of the geodesic is the path point of the required parallel equidistant path; fit the path points generated by all parallel equidistant paths into a spline curve.

[0021] Step 5: Take the spline curve generated in Step 4 as the new path reference, and repeat Step 4 until the laying path completely covers the surface to be laid; output the laying path, where the path points in the output path include the XYZ coordinate values and the Euler angles corresponding to the attitude of the laying head during laying. Convert the Euler angles into the quaternion form recognizable by the ABB robot to complete the path planning process;

[0022] Step 6: After the path output is completed, based on the automatic laying simulation platform, make the tool coordinate system, workpiece coordinate system coincide with the actual on-site coordinate system, import data for the simulation of the automatic laying path, detect the collision and interference problems between the laying head, positioner and die, and at the same time verify the correctness of the path points; after the path is correct and there is no interference between devices, the automatic laying experiment of composite materials can be carried out.

[0023] The beneficial effects of the present invention are as follows:

[0024] Through experiments, the present invention measures the bandwidth after the fiber placement under specific working conditions (temperature, pressure, speed, etc.), and feeds back this bandwidth to the path planning system, thereby reducing the gaps and overlaps during the automatic placement process and ensuring the quality of the component to be laid; compared with manual placement, in the face of complex curved surface components, the placement accuracy and quality are greatly improved, and at the same time the labor cost is reduced. Description of the Drawings

[0025] Figure 1 Schematic diagram for generating path points of fixed-angle iteration under triangular mesh;

[0026] Figure 2 Schematic diagram for generating the initial path under triangular mesh;

[0027] Figure 3 Schematic diagram of the parallel densification method based on the reference line;

[0028] Figure 4 Schematic diagram of the automatic laying path of the rotary body component, where (a) is the initial path, (b) is the path after parallel densification, and (c) is the path after introducing boundary conditions;

[0029] Figure 5 Schematic diagram of the automatic laying path simulation; Detailed Embodiments

[0030] In order to describe the invention purpose, invention advantages and technical solutions of the present invention more clearly, the following will describe the present invention in detail in combination with the drawings and implementation methods. However, it should be understood that the following described implementation methods are only used to explain the present invention and are not limited to the present invention. The present invention will be described in detail below. Through these detailed descriptions, those skilled in the art can fully understand the present invention.

[0031] Embodiment:

[0032] This embodiment provides a path generation and post - processing system for robotic automatic placement, whose goal is to plan the placement path according to the component design requirements, and to implement the post - processing and simulation of the placement path based on the RoboDK platform, and to complete the forming of composite components using an automatic placement equipment.

[0033] Step 1: The model definition module defines the geometric dimensions and structure of the surface to be placed by importing a model. The model file is an STL file. The STL file is a mesh model file obtained by minimizing geometric criteria, which approximates the CAD solid model with a large number of spatial triangular patches. By reading the STL file information, the information on each triangular patch unit is obtained, mainly including vertex coordinates, patch labels, normal vectors, etc.

[0034] Step 2: The initial path generation module defines the reference line (usually the axis), the path start point, and the ply angle of the placement path. Through the distance calculation formula:

[0035]

[0036] Traverse the triangular patch units to retrieve the triangular patch unit where the current start point P is located. Among them, D is the distance from point P to the triangular patch, n is the normal vector of the triangular patch, P is the path start point, and A and B are certain vertices on the triangular patch. When D = 0, the found triangular patch is obtained.

[0037] As Figure 1 shown, d is the direction of the reference line of the path, and calculate the projection d' of the reference line direction d under this triangular patch:

[0038]

[0039] Based on the projection direction d', calculate the actual placement direction v. The angle between v and d' is the set ply angle. The calculation formula of the rotation matrix RM is:

[0040]

[0041] Among them, I is the identity matrix, is the ply angle, and K is the skew - symmetric matrix of n:

[0042]

[0043] Thus, the placement direction v is:

[0044] v=(RM·d' T ) T

[0045] Solve the intersection point P of the line passing through point P along direction v and the boundary of the triangular patch P 1:

[0046] Establish the parametric equation: P(t) = P + t·v

[0047] Solve for the intersection point: LAB(s) = A + s(B - A), 0 ≤ s ≤ 1

[0048] Set up the equation: P + tv = A + s(B - A)

[0049] Solve the equation: Calculate the intersection point: P 1 = P + t·v.

[0050] Step 3: P 1 is the next laying point of the initial path. As Figure 2 shown, define P 1 as the starting point of the path for the next patch unit. Repeat Step 2 until the laying boundary is reached to obtain a series of path points {P 0 , P 1 , P 2 , P 3 ……}. Obtain the initial path by cubic B-spline fitting.

[0051] Step 4: When the process parameters of automatic laying are known, test the bandwidth w of the prepreg with a specified bandwidth after laying and forming through a laying experiment. Densify the initial path using the parallel equidistant method. The parallel equidistant principle is as Figure 3 shown. Discretize the initial path according to certain rules to obtain the binormal vector e 2 of the path at each discrete point. Take the above experimental result w as the offset distance. According to the distance w and the binormal vector direction e 2 generate geodesic segments with specified starting points, directions, and lengths. If the length of the geodesic segment is less than L, discard this position and execute the above steps at the next discrete point; if the length of the geodesic segment is equal to L, the end point of the geodesic is the path point of the required parallel equidistant path. Fit the path points generated by all parallel equidistant paths into a spline curve.

[0052] Step 5: Take the curve generated in Step 4 as the new path reference and repeat Step 4 until the laying path completely covers the surface to be laid. Output the laying path in the form of a csv file. The output path points include the XYZ coordinate values and the Euler angles of the laying head attitude during laying. Convert the Euler angles into the quaternion form recognizable by the ABB robot through a transformation matrix, thus ensuring that the laying head is always perpendicular to the surface of the surface to be laid, and then completing the path planning process.

[0053] Step 6: After the path output is completed, first build a composite material automatic placement simulation platform based on RoboDK, align the tool coordinates of the robotic arm, positioner, placement head, etc. with the workpiece coordinates and the actual on-site coordinate system, import the generated CSV data point file to realize the simulation of the automatic placement path, detect the collision and interference problems between the placement head, positioner and the mold, and at the same time verify the correctness of the path points. After the path is correct and there is no interference between the devices, the composite material automatic placement experiment can be carried out.

[0054] As Figure 4 shown, they are the initial path, densified path and the path diagram after introducing boundary conditions during the automatic placement path planning of a certain type of circular tube structure. After building the post-processing simulation platform through Step 6 and verifying that there is no collision and interference, it is output through the RAPID program and imported into the robotic arm equipment for the composite material automatic placement experiment.

[0055] The experimental results show that: the generated fiber placement trajectory has a good effect, the fiber bundle can evenly cover the surface of the mandrel, and the gap and overlap defects are greatly reduced, meeting the process design requirements and engineering needs.

[0056] The above detailed description of the present invention is only a preferred embodiment and is not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A robotic automatic placement path generation and post-processing simulation system, characterized in that: It includes model definition module, initial path generation module, path densification module and path post-processing simulation module; The model definition module is used to define the geometric model data of the surface to be laid; The initial path generation module is used to define the parameters of the laying path, calculate and generate path points in combination with the defined parameters, and fit the generated path points into the required initial path; The path densification module uses a parallel equidistance method to densify the initial path; The path post-processing simulation module outputs the curves planned in the densified path, automatically lays out the simulation platform, and performs automatic laying simulation, and then automatically lays out after the simulation verification is correct.

2. The robotic automatic placement path generation and post-processing simulation system according to claim 1, characterized in that: The initial path generation module calculates the triangular face unit where the path starting point is located in combination with the defined model parameters, calculates the projection of the reference direction under the corresponding triangular face, and calculates the intersection of the path starting point along the projection direction and the boundary of the corresponding triangular face unit; Define the calculated intersection point as the new starting point; The intersection calculation process is repeated until the generated path points exceed the defined boundary conditions; the generated set of path points is fitted into the required initial path through cubic B-spline.

3. The robotic automatic placement path generation and post-processing simulation system according to claim 2, characterized in that: The path densification module uses the parallel equidistant method to perform path densification, and experimentally measures the bandwidth of the laid tow under specific working conditions. Taking the initial path as a reference, the geodesic direction is solved through the initial path point along the tangent direction of the laying direction, and the point along the geodesic direction through the initial path point and the distance from the initial path to the experimental bandwidth length is solved. The path points of all parallel equidistant paths are fitted, and the curve is used as the new initial path to repeat the steps until the surface boundary.

4. The robotic automatic placement path generation and post-processing simulation system according to claim 3, characterized in that: The path post-processing simulation module outputs the path planned curve through a csv file, builds a composite material automatic placement simulation platform based on RoboDK, imports the generated csv data point file according to the built tools and workpiece coordinate system, and performs automatic placement simulation. After verification, the Rapid program can be output for automatic placement.

5. The simulation method of the robot-type automatic placement path generation and post-processing simulation system according to claim 4 is characterized in that: The following steps are involved: Step 1: The model definition module imports the model to define the geometric size and structure of the surface to be laid. The model file is an STL file. By reading the STL file information, the information on each triangular patch unit is obtained; Step 2: The initial path generation module defines the reference line, path starting point and ply angle of the laying path, traverses the triangular patch unit to retrieve the triangular patch unit where the current starting point P is located; calculates the projection direction d′ of the reference line direction d under the corresponding triangular patch, and uses the projection direction d′ as a reference to calculate the actual laying direction v, where the angle between v and d′ is the set ply angle; passes through point P and along the direction v to solve the intersection point P1 with the boundary of the triangular patch; Step 3: P1 is the next placement point of the initial path. P1 is defined as the path starting point of the next patch unit. Step 2 is repeated until the placement boundary is reached to obtain a set of path points {P0, P1, P2, P3...}. The initial path is obtained by cubic B-spline fitting. Step 4: When the process parameters of automatic placement are known, the bandwidth w of the prepreg with a specified bandwidth after placement is tested, and the initial path is densified using the parallel equidistant method; the initial path is discretized to obtain the binormal vector e2 of the path at each discrete point, and the tested bandwidth w is taken as the offset distance. According to the offset distance w and the binormal vector direction e2, a geodesic line segment with a specified starting point, direction and length is generated. When the length of the geodesic line segment is less than the set threshold value L, the step of generating the geodesic line segment is repeated at the next discrete point; when the length of the geodesic line segment is equal to the set threshold value L, the end point of the geodesic line is the path point of the parallel equidistant path required; the path points generated by all parallel equidistant paths are fitted into a spline curve; Step 5: Use the spline curve generated in step 4 as a new path reference, and repeat step 4 until the placement path completely covers the surface to be placed; output the placement path, and the path points in the output path include XYZ coordinate values ​​and the Euler angles corresponding to the placement head posture during placement. Convert the Euler angles into quaternion forms recognizable by the ABB robot to complete the path planning process; Step 6: After the path output is completed, based on the automatic placement simulation platform, the tool coordinates and workpiece coordinates are matched with the actual coordinate system on site, and the data is imported to simulate the automatic placement path, detect the collision and interference problems between the placement head, positioner and mold, and verify the correctness of the path points. After the path is correct and there is no interference between the equipment, the automatic placement experiment of the composite material can be carried out.

Images