A trajectory optimization method for laser welding
Patent Information
- Application Number
- CN202610754854.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-28
- Publication Date
- 2026-08-21
AI Technical Summary
该问题会影响焊接热输入分布和光斑作用稳定性，导致轨迹满足位置连续要求时仍难以保证焊接过程的姿态连续性
本发明通过基于焊缝局部几何建立接头截面，将机器人逆运动学解对应的光束轴线投影至接头截面以获得入射角变化方向，并利用姿态雅可比矩阵将关节角增量映射为末端姿态变化方向，再依据二者相对焊缝中心线的方向符号确定机器人逆运动学解配对，进而生成覆盖各轨迹点的机器人逆运动学解序列，使轨迹优化过程同时考虑TCP位置连续性和焊接头姿态解连续性，降低姿态解分支变化引起的光束入射关系跳变，提升空间焊缝激光焊接过程中的光斑作用稳定性。
Smart Images
Figure CN122606148A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of laser welding technology, and more specifically to a trajectory optimization method for laser welding. Background Technology
[0002] Laser welding trajectory planning in automated welding equipment is mainly used to control the laser's point of application along a predetermined weld seam and maintain the spatial orientation relationship between the weld head and the joint area. For weld seams with significant spatial variations, the trajectory execution process is affected not only by the continuity of the path position but also by the changes in the robot's posture. If the execution trajectory is generated solely based on the weld seam's geometric path, the position trajectory may easily meet the following requirements, while the beam incidence state may exhibit discontinuous changes between adjacent trajectory points.
[0003] Existing trajectory planning methods typically focus on weld path fitting and robot reachability assessment, lacking collaborative constraints on the continuous relationships between multiple attitude solutions at the same trajectory point. When the robot controller experiences attitude solution branching changes during execution, the welding head attitude changes accordingly, causing a jump in the incident relationship of the laser beam relative to the joint area. This problem affects the welding heat input distribution and the stability of the laser spot, making it difficult to guarantee the attitude continuity of the welding process even when the trajectory meets the positional continuity requirements. Summary of the Invention
[0004] The purpose of this invention is to provide a trajectory optimization method for laser welding to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a trajectory optimization method for laser welding, comprising: The joint section is established based on the welding trajectory execution data containing the local geometry of the weld, and the robot inverse kinematics solution for each trajectory point is obtained based on the robot kinematics model; The beam axis corresponding to the welding head pose is projected onto the joint section by the robot's inverse kinematics solution to obtain the direction of the incident angle change; Using the joint angles corresponding to the robot's inverse kinematics solution as the points for calculating the posture Jacobian matrix, the joint angle increments of adjacent trajectory point pairs are mapped to the end-effector posture increments. The end-effector posture increments are applied to the beam axis and projected onto the joint section to obtain the end-effector posture change direction. The robot's inverse kinematics solution pairing is determined according to the direction signs of the incident angle change direction and the end-effector posture change direction relative to the weld centerline. Based on the robot inverse kinematics solution pairing, a sequence of robot inverse kinematics solutions covering each trajectory point is generated, and the laser welding optimization trajectory is generated from the robot inverse kinematics solution sequence.
[0006] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention establishes a joint cross-section based on the local geometry of the weld, projects the beam axis corresponding to the robot's inverse kinematics solution onto the joint cross-section to obtain the direction of incident angle change, and uses the attitude Jacobian matrix to map the joint angle increment to the end-effector attitude change direction. Then, based on the direction signs of the two relative to the weld centerline, the robot's inverse kinematics solution pairing is determined, thereby generating a sequence of robot inverse kinematics solutions covering each trajectory point. This allows the trajectory optimization process to simultaneously consider the continuity of TCP position and the continuity of the weld head attitude solution, reducing the jump in beam incident relationship caused by attitude solution branch changes and improving the stability of the beam spot effect during the laser welding process of spatial welds. Attached Figure Description
[0007] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.
[0008] Figure 1 This is a flowchart of a trajectory optimization method for laser welding according to the present invention. Detailed Implementation
[0009] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make the description of this application more complete and comprehensive, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are merely illustrative illustrations of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.
[0010] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of the exemplary embodiments disclosed in this application. However, those skilled in the art will recognize that the technical solutions disclosed in this application can be practiced with one or more specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the disclosure of this application.
[0011] Example 1 like Figure 1 As shown, this embodiment discloses a trajectory optimization method for laser welding, including: S101: Establish the joint section based on the welding trajectory execution data containing the local geometry of the weld, and obtain the robot inverse kinematics solution for each trajectory point based on the robot kinematics model; The welding trajectory execution data is a data set arranged by trajectory point number, and the trajectory point number corresponds to the TCP position, local geometry of the weld, and robot inverse kinematics solution inputs.
[0012] Welding trajectory execution data is organized according to the welding execution sequence. Trajectory point numbers are denoted as follows: The trajectory point numbers are arranged sequentially along the weld seam extension direction. The TCP position corresponding to each trajectory point is denoted as . The corresponding weld center position is denoted as The TCP position is the tool center point position in the robot's execution trajectory. The weld center position is the spatial point representing the local center of the weld within the metal joint area. The local geometry of the weld includes the weld center position, the joint surface normal, the reference direction on the joint opening side, and the local geometric relationships of the two surfaces of the joint. The input terms for solving the robot's inverse kinematics include the robot link parameters, joint range of motion, and the pose transformation relationship between the robot's end effector coordinate system and the welding head tool coordinate system.
[0013] When the TCP position coincides with the weld center position, the weld center position Take as TCP position When the laser welding process requires a process offset between the TCP position and the weld center position, the TCP position... Used for solving robot inverse kinematics, weld center position. Used to establish the joint cross-section. Through the above data differentiation method, the robot's execution position and the joint's geometric center participate in corresponding processing respectively, so that the subsequent joint cross-section establishment, beam axis projection, and direction sign calculation all have a consistent geometric reference.
[0014] Specifically, the process of establishing the joint cross-section includes: Read the weld center position corresponding to the trajectory point in the welding trajectory execution data, calculate the spatial change of the weld center position along the trajectory point sequence, and obtain the weld tangent vector; Read the welding trajectory execution data. The weld center position corresponding to each trajectory point For non-end trajectory points, read the weld center position of the previous trajectory point. The weld center position of the next trajectory point The weld tangent vector is calculated based on the spatial variation of the weld center position along the trajectory points. :
[0015] In the formula, Indicates the first The weld tangent vector at each trajectory point Let the vector norm be denoted. For the first trajectory point, the weld tangent vector... Calculate using the following formula:
[0016] For the last trajectory point, the weld tangent vector Calculate using the following formula:
[0017] In the formula, This represents the total number of trajectory points. The weld tangent vector characterizes the local extension direction of the weld centerline at each trajectory point. For spatial welds or welds on bent metal parts, the extension direction of the weld centerline varies with the trajectory points. Calculating the weld tangent vector based on the center positions of adjacent welds allows the cross-sectional orientation of the joint section to change synchronously with the local weld orientation, providing a local geometric reference for the subsequent projection of the beam axis into the joint section.
[0018] The joint normal vector is obtained based on the spatial relationship between the joint surface normal and the weld tangent vector in the local geometry of the weld. Read the first The surface normal vectors of the two sides of the joint at each trajectory point are denoted as follows: and The surface normal vector lies in the robot's base coordinate system. If the surface normal vector originates from the workpiece model coordinate system, it is first transformed to the robot's base coordinate system based on the workpiece calibration matrix before being used in the joint normal vector calculation.
[0019] Before participating in the joint normal vector calculation, the surface normal vectors on both sides of the joint are oriented correctly. Orientation correction is based on the reference direction on the joint opening side in the local weld geometry, ensuring that the corrected surface normal vectors all point towards the joint opening side or the outside of the fusion zone. The corrected surface normal vectors are still denoted as... and This process ensures that the normal vectors of both surfaces have a uniform orientation when calculated along the bisector of the included angle.
[0020] Project the surface normal vectors of both sides of the joint onto the perpendicular tangent vector of the weld. The plane is used to obtain the surface normal vector after projection. and :
[0021]
[0022] In the formula, This represents the dot product of vectors. The projected surface normal vector is normalized to obtain the normalized surface normal vector. and :
[0023]
[0024] Calculate the joint normal vector along the bisector of the angle between the normalized surface normal vectors on both sides. :
[0025] In the formula, Indicates the first The joint normal vector at each trajectory point. The joint normal vector is jointly determined by the normals of the two surfaces of the joint and the weld tangent vector, and is used to characterize the geometric direction within the joint cross-section from the weld centerline to the joint opening side or the main direction of the fusion region. This joint normal vector does not depend on the normal direction of a single surface and can be applied to joint types such as corner joints, lap joints, and bent edges.
[0026] The direction of the joint section is determined by the weld tangent vector, and the normal side of the weld centerline within the joint section is determined by the joint normal vector, thus establishing the joint section.
[0027] Using weld tangent vector As the normal direction of the joint cross section, with the weld center position As the spatial point through which the joint cross-section passes, establish the first The joint cross section corresponding to each trajectory point :
[0028] In the formula, Indicates the joint cross-section Any spatial point within the area. Joint cross-section. Perpendicular to the weld tangent vector , used to express the first The cross-section of the joint at each trajectory point, perpendicular to the local extension direction of the weld.
[0029] At the joint cross section Inside, at the center of the weld The point where the weld centerline intersects the joint cross-section. Joint normal vector. It has been obtained from the plane projection perpendicular to the weld tangent vector, and is therefore located at the joint section. Inside. Using the joint normal vector. The side pointing to the center line serves as the normal side of the weld centerline. The joint cross-section, the weld centerline, and the normal side of the weld centerline together constitute the geometric reference for subsequent beam cross-section projection, incident angle change direction calculation, and direction sign calculation.
[0030] After establishing the joint cross-section, the inverse kinematics solutions for each trajectory point are obtained based on the robot's kinematic model. The robot kinematic model includes the robot base coordinate system, robot joint coordinate system, robot end effector coordinate system, link dimension parameters, joint type, joint range of motion, and forward kinematic mapping relationships. For the... Each trajectory point is determined based on the TCP position in the welding trajectory execution data. Constructing the robot's end-effector pose using welding head posture constraints The orientation constraints of the welding head are determined by the orientation transformation relationship between the welding head tool coordinate system and the robot end effector coordinate system, as well as the initial incident relationship between the beam axis and the joint cross section.
[0031] Position the robot's end effector Input the robot's kinematics model and perform inverse kinematics solution to obtain the... The set of robot inverse kinematics solutions corresponding to each trajectory point Robot inverse kinematics solution set Including enabling the robot's end effector to reach One or more joint angle vectors. At the i-th trajectory point The inverse kinematics solution of the robot is denoted as For a six-degree-of-freedom robot, Represented as:
[0032] In the formula, Indicates the first At the i-th trajectory point The first inverse kinematics solution of the group robot Each joint angle. Multiple sets of robot inverse kinematics solutions can exist for the same trajectory point. Different robot inverse kinematics solutions may correspond to the same TCP position, but to different joint space configurations and different welding joint posture change trends. Therefore, the output of S101 includes the joint cross-section corresponding to each trajectory point. and the solution set of robot inverse kinematics The aforementioned output provides the data foundation for subsequent beam axis projection and robot inverse kinematics pairing determination.
[0033] It should be noted that the joint section establishment in S101 belongs to the geometric reference construction process, while the robot inverse kinematics solution acquisition belongs to the robot joint space solution process. The two are established with the same trajectory point number as an index, enabling subsequent processing to simultaneously call the joint section and the robot inverse kinematics solution at the same trajectory point. This correspondence helps avoid ignoring the continuity of the weld joint attitude solution when trajectory planning is performed solely based on the TCP position.
[0034] S102: Project the beam axis corresponding to the welding head pose of the robot's inverse kinematics solution onto the joint section to obtain the direction of the incident angle change; For each set of robot inverse kinematics solutions at each trajectory point, the robot end effector pose corresponding to that inverse kinematics solution is calculated, and the welding head pose is obtained based on the pose transformation relationship between the robot end effector pose and the welding head tool coordinate system. The welding head pose is used to determine the spatial orientation of the beam axis in the robot base coordinate system. Subsequently, the beam axis is projected onto the joint cross-section of the corresponding trajectory point to form the beam cross-section projection line. By comparing the angular change direction of the beam cross-section projection line relative to the weld centerline of adjacent trajectory point pairs, the incident angle change direction is obtained.
[0035] Specifically, the process of obtaining the direction of the incident angle change includes: The robot end-effector pose is calculated by solving the robot inverse kinematics, and the welding head pose is obtained by the pose transformation relationship between the robot end-effector pose and the welding head tool coordinate system. For the At the i-th trajectory point Inverse kinematics solution of group robots Calculate the robot's end-effector pose based on the robot's forward kinematics model. Robot end effector pose Let be the homogeneous transformation matrix of the robot end-effector coordinate system relative to the robot base coordinate system. Let be the pose transformation matrix of the welding head tool coordinate system relative to the robot end-effector coordinate system. Then the first At the i-th trajectory point The inverse kinematics solution of the robot corresponds to the welding head pose. for:
[0036] Will The rotation matrix in the matrix is denoted as The laser emission axis in the welding head tool coordinate system is denoted as the unit direction vector. .Will The beam axis direction vector is obtained by transforming the welding head tool coordinate system to the robot base coordinate system. :
[0037] In the formula, Indicates the first At the i-th trajectory point The beam axis direction vector corresponds to the robot inverse kinematics solution. Since the beam axis direction vector is jointly determined by the robot inverse kinematics solution, the robot end effector pose, and the welding head tool coordinate system, different robot inverse kinematics solutions at the same TCP position can correspond to different beam axis directions. By establishing a correspondence between the beam axis and the robot inverse kinematics solution, the influence of the attitude solution selection on the beam incident direction can be incorporated into subsequent calculations.
[0038] The beam axis of the welding head is projected onto the joint cross-section to form the beam cross-section projection line; beam axis direction vector Orthogonal projection to the 1st The joint cross section corresponding to each trajectory point Joint cross-section The normal direction is the weld tangent vector. Therefore, the beam axis direction vector Projected direction vector within the joint cross section for:
[0039] For the projection direction vector After normalization, the unit projection direction vector is obtained. :
[0040] Center position of weld As the intersection point through which the beam cross-section projection line passes, As the direction of the beam section projection line, the beam section projection line is formed. :
[0041] In the formula, Indicates the first At the i-th trajectory point The beam section projection line corresponding to the inverse kinematics solution of the robot group. This represents the linear parameters. The beam cross-section projection line is used to characterize the incident direction of the beam axis relative to the weld centerline within the joint cross-section.
[0042] The process of forming the projection lines of the beam cross section includes: The direction vector of the beam axis is orthogonally projected onto the joint section to obtain the projected direction vector within the joint section; Based on the normal direction of the joint section Remove beam axis direction vector along The component, retained at the joint cross-section The components within the vector are used to obtain the projection direction vector. This projection direction vector converts the spatial direction of the beam axis into a direction within the joint cross-section, providing a unified geometric reference for subsequent calculations of angular change directions.
[0043] The projection direction vector passes through the intersection point of the weld centerline in the joint section, forming the beam section projection line.
[0044] Center position of weld As the centerline of the weld at the joint cross section The intercept point in the vector makes the projection direction vector At this intercept point, the beam cross-section projection line is formed. The beam cross-section projection lines corresponding to different robot inverse kinematics solutions all take the weld center position at the same trajectory point as the intercept point. Therefore, the angular difference between different beam cross-section projection lines can characterize the influence of different robot inverse kinematics solutions on the beam incident relationship.
[0045] By comparing the angle change direction of the beam cross-section projection line relative to the weld centerline of adjacent trajectory point pairs, the direction of incident angle change can be obtained.
[0046] The first The trajectory point and the first The trajectory points form adjacent trajectory point pairs. For the , At the i-th trajectory point Inverse kinematics solution of the group robot and the first At the i-th trajectory point The inverse kinematics solution of the group of robots was used to obtain the beam cross-section projection lines. and .Will relative joint normal vector The directed angle is denoted as ,Will relative joint normal vector The directed angle is denoted as .in:
[0047]
[0048] In the formula, Represents the arctangent function with quadrant recognition. Represents the cross product of vectors. and These represent the unit projection direction vectors of the beam axis at adjacent trajectory points within the corresponding connector cross-section. The aforementioned directional angle represents the angle from the connector normal vector within the corresponding connector cross-section to the projection direction of the beam cross-section.
[0049] The angle change of adjacent trajectory point pairs is:
[0050] according to Determine the direction of change of the incident angle relative to the direction normal to the weld centerline. If The direction of change in the incident angle is denoted as the positive direction; if The direction of change in the incident angle is recorded as negative; if The direction of change of the incident angle is denoted as the zero change direction.
[0051] Through the above processing, the beam axis corresponding to each set of robot inverse kinematics solutions is mapped to the beam cross-section projection line within the joint cross-section. The beam cross-section projection lines of adjacent trajectory point pairs are further converted into the incident angle variation direction. This incident angle variation direction is used to characterize the incident change trend of the laser beam axis along the trajectory point sequence within the joint cross-section, and serves as the geometric basis in subsequent robot inverse kinematics solution pairing judgment. This processing is beneficial for identifying robot inverse kinematics solution combinations where the TCP position meets the weld seam following requirements but the beam incident direction undergoes cross-side changes.
[0052] S103: Using the joint angles corresponding to the robot inverse kinematics solution as the points for calculating the posture Jacobian matrix, the joint angle increments of adjacent trajectory point pairs are mapped to the end-effector posture increments; the projection of the beam axis in the joint section is updated by the end-effector posture increments to obtain the end-effector posture change direction; the robot inverse kinematics solution pairing is determined according to the direction sign of the incident angle change direction and the end-effector posture change direction relative to the weld centerline. Wherein, the attitude Jacobian matrix is the attitude mapping matrix in the robot Jacobian matrix, the matrix rows of the attitude mapping matrix correspond to the end attitude angular velocity components, and the matrix columns of the attitude mapping matrix correspond to the joint angular velocity components.
[0053] No. The trajectory point and the first Each trajectory point forms a pair of adjacent trajectory points. At the i-th trajectory point The inverse kinematics solution of the robot is denoted as , No. At the i-th trajectory point The inverse kinematics solution of the robot is denoted as These two elements form a candidate robot inverse kinematics solution pair, used to calculate joint angle increments, end-effector pose increments, and end-effector pose change directions. Through this pairing relationship, a correspondence is established between joint space changes and beam incident direction changes between adjacent trajectory points.
[0054] The attitude Jacobian matrix is taken from the angular velocity mapping part of the robot's Jacobian matrix. For a six-DOF robot, the attitude Jacobian matrix is expressed as follows: ,in, Let be the robot joint angle vector. Using the joint angles corresponding to the robot's inverse kinematics solution as the evaluation points, the pose Jacobian matrix reflects the influence of the joint angle increment on the robot's end-effector pose increment under the current joint configuration.
[0055] Specifically, the process of obtaining the end-effector attitude change direction includes: For the robot inverse kinematics solutions in adjacent trajectory point pairs, joint angle correspondence is performed to obtain the joint angle increment; For the At the i-th trajectory point Inverse kinematics solution of group robots and the At the i-th trajectory point Inverse kinematics solution of group robots Establish the joint angle correspondence according to the robot joint number. For a six-degree-of-freedom robot, the two are represented as follows:
[0056]
[0057] In the formula, Indicates the first At the i-th trajectory point The first inverse kinematics solution of the group robot One joint angle, Indicates the first At the i-th trajectory point The first inverse kinematics solution of the group robot Each joint angle is assigned a number of joint angles. The joint angle correspondence is established according to the same joint number to ensure that the joint angle increment reflects the changes of the same robot joint between adjacent trajectory points.
[0058] The calculation process for the joint angle increment includes: According to the correspondence of robot joint numbers, the joint angles corresponding to the robot inverse kinematics solutions in adjacent trajectory point pairs are differentially analyzed item by item to obtain the original joint angle difference values. Based on the correspondence of robot joint numbers, the joint angles in the inverse kinematics solution pairing of candidate robots are subtracted term by term to obtain the original joint angle difference values. :
[0059] In the formula, Indicates the pairing of candidate robot inverse kinematics solutions The Middle The original joint angle difference values of each joint. The original joint angle difference values of all joints form the original joint angle difference value vector. :
[0060] The original joint angle difference is used to represent the change in joint angle between adjacent trajectory points. For rotary joints, it is also necessary to perform unfolding processing in conjunction with the joint rotation period to ensure that the joint angle increment corresponds to the actual rotation direction.
[0061] The original joint angle difference is processed by joint rotation period expansion to obtain the joint angle increment.
[0062] For a rotary joint, the rotation period is denoted as . For the original joint angle difference value Perform joint rotation cycle unfolding processing to obtain joint angle increments. :
[0063] In the formula, This indicates a floor operation. After the above processing, the joint angle increment of the rotary joint is... lie in Within the range. For moving joints, the joint angle increment is taken as the difference in joint displacement. The joint angle increments of all joints form the joint angle increment vector. :
[0064] The joint rotation period expansion processing ensures that the joint angle increment corresponds to the actual joint movement direction, avoiding discontinuous representation of joint angle differences caused by periodic expression for the same rotation joint. This processing provides continuous joint space input for subsequent pose Jacobian matrix mapping.
[0065] The posture Jacobian matrix is calculated by taking the joint angle corresponding to the robot inverse kinematics solution of the previous trajectory point in the adjacent trajectory point pair as the evaluation point; With the first At the i-th trajectory point Inverse kinematics solution of group robots To find the evaluation point, the pose Jacobian matrix is calculated based on the robot's kinematic model. The attitude Jacobian matrix is the matrix part of the robot's Jacobian matrix that describes the mapping relationship between the end-effector angular velocity and the joint angular velocity. For a six-DOF robot, the attitude Jacobian matrix is expressed as:
[0066] In the formula, Indicates at the joint angle The pose Jacobian matrix is obtained at the given point. This matrix is jointly determined by the robot link parameters, joint axis directions, and the current joint angle. Using the joint angle corresponding to the robot's inverse kinematics solution of the previous trajectory point as the calculation point, the local pose mapping relationship of the robot when moving from the previous trajectory point to the next trajectory point can be characterized.
[0067] The joint angle increment is input into the attitude Jacobian matrix to obtain the end attitude increment. The end attitude increment is then applied to the beam axis and projected onto the joint section to obtain the end attitude change direction.
[0068] Joint angle increment vector Input attitude Jacobian matrix Obtain the end attitude increment :
[0069] In the formula, Indicates the pairing of candidate robot inverse kinematics solutions The end-effector attitude increment corresponds to the joint angle increment. The end-effector attitude increment is represented by a rotation vector in the robot's base coordinate system.
[0070] To update the beam axis based on the end-effector attitude increment, a structure is constructed using... The resulting antisymmetric matrix .when hour:
[0071] Based on the exponential mapping of the rotation vector, the rotation matrix corresponding to the end-effector attitude increment is obtained:
[0072] Apply the rotation matrix to the beam axis direction vector at the previous trajectory point The updated beam axis direction vector is obtained. :
[0073] Then Projected onto the joint cross section Obtain the updated projection direction vector :
[0074] The projection direction vector of the original beam axis at the previous trajectory point onto the joint section is: The direction of end-effector attitude change is obtained based on the change between the updated projection direction vector and the original projection direction vector. :
[0075] In the formula, Indicates the direction of the end-effector attitude change. If If the direction sign of the end effector's attitude change direction is zero, then the change in the robot's joint space is converted into the change direction of the beam projection within the joint cross-section, which is under the same geometric reference as the change direction of the incident angle obtained in S102. This processing is helpful in distinguishing between the incident change caused by the local geometric changes of the weld and the attitude change caused by the switching of the robot's inverse kinematics solution.
[0076] Specifically, the process of determining the robot's inverse kinematics pairing includes: Combine any robot inverse kinematics solution of the preceding trajectory point with any robot inverse kinematics solution of the following trajectory point in an adjacent trajectory point pair to form a candidate robot inverse kinematics solution pair. For adjacent trajectory point pairs , No. The set of robot inverse kinematics solutions for each trajectory point is as follows: , No. The set of robot inverse kinematics solutions for each trajectory point is as follows: Choose any and Pairing of candidate robot inverse kinematics solutions:
[0077] The candidate pairing set is formed by pairing all candidate robot inverse kinematics solutions. :
[0078] In the formula, Indicates the first The trajectory point and the first The candidate pairing set between trajectory points. The candidate pairing set covers the robot inverse kinematics solutions that can be used for judgment between adjacent trajectory points, providing input for subsequent direction sign calculation and pairing determination.
[0079] Calculate the direction signs of the incident angle change direction and the end attitude change direction along the normal side of the weld centerline in the joint section, and obtain the incident angle direction sign and the end attitude direction sign. For candidate robot inverse kinematics solution pairing The direction of change of the incident angle is determined by the amount of change in angle. Characterization. Incident angle direction sign. Obtained by the following formula:
[0080] In the formula, Represents a symbolic function. When... hour, ;when hour, ;when hour, .
[0081] End attitude change direction Located within the joint cross-section. Project the direction of the end-position change onto the normal reference direction of the weld centerline. Obtain the end attitude direction sign :
[0082] In the formula, This represents the end-effector attitude direction symbol corresponding to the inverse kinematics solution pair of the candidate robot. Through this method, both the incident angle change direction and the end-effector attitude change direction are converted into direction symbols relative to the normal side of the weld centerline, and both are under the same joint section reference.
[0083] Candidate robot inverse kinematics solutions with the same incident angle direction sign and end-effector attitude direction sign are paired and identified as robot inverse kinematics solution pairs.
[0084] For candidate robot inverse kinematics solution pairing When the following conditions are met: When a candidate robot inverse kinematics solution pairing is selected, it is considered a robot inverse kinematics solution pairing. Candidate robot inverse kinematics solution pairs with different direction signs are not considered robot inverse kinematics solution pairs. Robot inverse kinematics solution pairs that satisfy the condition of having the same direction sign are grouped into a pairing set. :
[0085] In the formula, Indicates the first The trajectory point and the first The robot inverse kinematics solutions are paired up between trajectory points. This process unifies the direction of incident angle change and end-effector attitude change into a single direction sign determination within the junction section. This approach helps reduce the impact of attitude solution switching on the continuity of beam incidence under the same TCP position trajectory.
[0086] S104: Generate a sequence of robot inverse kinematics solutions covering each trajectory point based on the robot inverse kinematics solution pairing, and generate the laser welding optimized trajectory from the robot inverse kinematics solution sequence.
[0087] Robot inverse kinematics solution pairing represents the connectability between adjacent trajectory points. To form a complete welding trajectory, robot inverse kinematics solution pairs need to be connected sequentially along the trajectory points to obtain a sequence of robot inverse kinematics solutions covering all trajectory points. Each item in the robot inverse kinematics solution sequence corresponds to the robot inverse kinematics solution of a trajectory point, and adjacent items belong to the robot inverse kinematics solution pairing between corresponding trajectory points.
[0088] Specifically, the process of generating the optimized laser welding trajectory includes: Connect the robot inverse kinematics solutions sequentially along the trajectory points to generate a sequence of robot inverse kinematics solutions covering each trajectory point; Inverse kinematics solution for each robot As nodes, the robot's inverse kinematics is unpaired. This represents the directed connection between adjacent trajectory points. The robot inverse kinematics solution set is derived sequentially from the first trajectory point. Begin by sequentially connecting the pairs belonging to each set. Pairing of robot inverse kinematics solutions. If a sequence of robot inverse kinematics solutions exists. :
[0089] And for any All satisfy:
[0090] Then the robot inverse kinematics solution sequence Covers all trajectory points. In the formula, Indicates the first The sequence number of the robot inverse kinematics solution selected for each trajectory point.
[0091] When multiple robot inverse kinematics solution sequences covering all trajectory points exist, they are sorted according to the cumulative joint angle increment. (Cumulative joint angle increment) Represented as:
[0092] according to The values are sorted from smallest to largest, and the robot inverse kinematics solution sequence with the first value is taken as the robot inverse kinematics solution sequence used to generate robot joint commands.
[0093] When no robot inverse kinematics solution sequence covering all trajectory points exists, the existing robot inverse kinematics solution sequence covering the continuous trajectory point number range is retained, and the uncovered trajectory points are designated as trajectory points requiring regeneration of robot inverse kinematics solutions. For uncovered trajectory points, the TCP position, welding joint attitude constraints, and robot kinematics model of the corresponding trajectory point are re-invoked to obtain a new set of robot inverse kinematics solutions, and the robot inverse kinematics solution pairings are formed according to the joint cross-section, beam axis projection, and direction sign judgment rules. This process ensures that the trajectory generation process still has an executable processing path when local candidate solutions are insufficient.
[0094] The robot's inverse kinematics solution sequence is converted into robot joint commands, and the robot joint commands are matched with the TCP positions in the welding trajectory execution data to generate an optimized laser welding trajectory.
[0095] Decompose the robot's inverse kinematics sequence Each robot inverse kinematics solution is converted into robot joint commands for the corresponding trajectory point. The robot joint commands for each trajectory point are denoted as follows: :
[0096] Robot joint commands TCP position in welding trajectory execution data Based on the corresponding trajectory point numbers, an optimized laser welding trajectory is formed. :
[0097] In the formula, This represents the optimized laser welding trajectory. The optimized laser welding trajectory includes TCP positions arranged in sequence according to the trajectory points and the corresponding robot joint commands. When the robot executes this optimized laser welding trajectory, the TCP positions move along the weld path in the welding trajectory execution data, and the robot joint commands are determined by a sequence of robot inverse kinematics solutions that satisfy the direction sign consistency relationship.
[0098] For example, for three consecutive trajectory points , and ,like ,and This allows for the formation of a sequence of local robot inverse kinematics solutions. Because adjacent pairs all use [a certain method] at intermediate trajectory points. This local sequence maintains consistency in the robot's inverse kinematics solution at intermediate trajectory points. This connection method reduces attitude switching caused by adjacent pairs using different robot inverse kinematics solutions at the same trajectory point.
[0099] Through processing in S104, the robot inverse kinematics solutions are paired and organized into a sequence of robot inverse kinematics solutions covering each trajectory point, and the laser welding optimized trajectory is generated from this sequence. This processing helps to extend the consistency of direction signs between adjacent trajectory points to the entire weld seam trajectory, ensuring that the TCP position trajectory and the welding head posture changes remain consistent during robot execution.
[0100] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A trajectory optimization method for laser welding, characterized in that, include: The joint section is established based on the welding trajectory execution data containing the local geometry of the weld, and the robot inverse kinematics solution for each trajectory point is obtained based on the robot kinematics model; The beam axis corresponding to the welding head pose is projected onto the joint section by the robot's inverse kinematics solution to obtain the direction of the incident angle change; Using the joint angles corresponding to the robot's inverse kinematics solution as the points for calculating the posture Jacobian matrix, the joint angle increments of adjacent trajectory point pairs are mapped to the end-effector posture increments. The end-effector posture increments are applied to the beam axis and projected onto the joint section to obtain the end-effector posture change direction. The robot's inverse kinematics solution pairing is determined according to the direction signs of the incident angle change direction and the end-effector posture change direction relative to the weld centerline. Based on the robot inverse kinematics solution pairing, a sequence of robot inverse kinematics solutions covering each trajectory point is generated, and the laser welding optimization trajectory is generated from the robot inverse kinematics solution sequence.
2. The method according to claim 1, characterized in that, The welding trajectory execution data is a data set arranged by trajectory point number, and the trajectory point number corresponds to the TCP position, local geometry of the weld, and robot inverse kinematics solution inputs.
3. The method according to claim 1, characterized in that, The process of establishing the joint cross-section includes: Read the weld center position corresponding to the trajectory point in the welding trajectory execution data, calculate the spatial change of the weld center position along the trajectory point sequence, and obtain the weld tangent vector; The joint normal vector is obtained based on the spatial relationship between the joint surface normal and the weld tangent vector in the local geometry of the weld. The direction of the joint section is determined by the weld tangent vector, and the normal side of the weld centerline within the joint section is determined by the joint normal vector, thus establishing the joint section.
4. The method according to claim 1, characterized in that, The process of obtaining the direction of the incident angle change includes: The robot end-effector pose is calculated by solving the robot inverse kinematics, and the welding head pose is obtained by the pose transformation relationship between the robot end-effector pose and the welding head tool coordinate system. The beam axis of the welding head is projected onto the joint cross-section to form the beam cross-section projection line; By comparing the angle change direction of the beam cross-section projection line relative to the weld centerline of adjacent trajectory point pairs, the direction of incident angle change can be obtained.
5. The method according to claim 4, characterized in that, The formation process of the beam cross-section projection line includes: The direction vector of the beam axis is orthogonally projected onto the joint section to obtain the projected direction vector within the joint section. The projection direction vector passes through the intersection point of the weld centerline in the joint section, forming the beam section projection line.
6. The method according to claim 1, characterized in that, The attitude Jacobian matrix is the attitude mapping matrix in the robot Jacobian matrix. The matrix rows of the attitude mapping matrix correspond to the end-effector angular velocity components, and the matrix columns of the attitude mapping matrix correspond to the joint angular velocity components.
7. The method according to claim 1, characterized in that, The process of obtaining the end-effector attitude change direction includes: For the robot inverse kinematics solutions in adjacent trajectory point pairs, joint angle correspondence is performed to obtain the joint angle increment; The posture Jacobian matrix is calculated by taking the joint angle corresponding to the robot inverse kinematics solution of the previous trajectory point in the adjacent trajectory point pair as the evaluation point; The joint angle increment is input into the attitude Jacobian matrix to obtain the end attitude increment. The end attitude increment is then applied to the beam axis and projected onto the joint section to obtain the end attitude change direction.
8. The method according to claim 7, characterized in that, The calculation process for the joint angle increment includes: Based on the correspondence of robot joint numbers, the joint angles corresponding to the robot inverse kinematics solutions in adjacent trajectory point pairs are differentially analyzed item by item to obtain the original joint angle difference values. The original joint angle difference is processed by joint rotation period expansion to obtain the joint angle increment.
9. The method according to claim 1, characterized in that, The process of determining the robot's inverse kinematics pairing includes: Combine any robot inverse kinematics solution of the preceding trajectory point with any robot inverse kinematics solution of the following trajectory point in an adjacent trajectory point pair to form a candidate robot inverse kinematics solution pair. Calculate the direction signs of the incident angle change direction and the end attitude change direction along the normal side of the weld centerline in the joint section, and obtain the incident angle direction sign and the end attitude direction sign. Candidate robot inverse kinematics solutions with the same incident angle direction sign and end-effector attitude direction sign are paired and identified as robot inverse kinematics solution pairs.
10. The method according to claim 1, characterized in that, The process of generating the laser welding optimized trajectory includes: Connect the robot inverse kinematics solutions sequentially along the trajectory points to generate a sequence of robot inverse kinematics solutions covering each trajectory point; The robot inverse kinematics solution sequence is converted into robot joint commands, and the robot joint commands are matched with the TCP positions in the welding trajectory execution data to generate an optimized laser welding trajectory.