Automated laser cleaning methods, systems, and media based on 3D graphics

Abstract

This application provides an automated laser cleaning method, system, and medium based on 3D graphics. The method includes: constructing a digital twin of the production line in a three-dimensional virtual environment, importing a workpiece model or scanning the workpiece to obtain a three-dimensional model of the workpiece, and configuring the characteristic parameters of the laser cleaning head; generating a cleaning path using 3D graphics algorithms; generating a robot motion trajectory using a gradient-based obstacle avoidance algorithm based on the interference obstacle avoidance algorithm, the cleaning path, and the robot kinematics model, and performing visual editing and program optimization; simulating cleaning in a three-dimensional virtual environment based on the physical principles of the laser cleaning head; executing a program simulation based on the cleaning simulation, collecting cleaning process data in real time, and storing the cleaning process data back to the virtual environment; and automatically generating the optimal cleaning path based on the workpiece's appearance characteristics through 3D graphics algorithms and laser cleaning process knowledge, thereby improving cleaning quality and efficiency.

Description

Technical Field

[0001] This application relates to the field of laser cleaning technology, and more specifically, to an automated laser cleaning method, system, and medium based on 3D graphics. Background Technology

[0002] With the rapid development of intelligent manufacturing and Industry 4.0, laser cleaning technology, as a highly efficient and environmentally friendly surface treatment technology, has been widely used in automobile manufacturing, aerospace, and machining. However, traditional laser cleaning systems face the following problems when dealing with complex workpieces and changing production demands: Disadvantages of existing technology: Inaccurate cleaning path planning: Existing systems lack intelligent path planning capabilities based on workpiece appearance features. Cleaning paths often rely on manual experience or simple geometric algorithms, failing to fully consider process parameters such as the optimal working range and incident angle of the laser cleaning head, resulting in unstable cleaning effects and low efficiency.

[0003] Obstacle avoidance algorithms are simplistic and inflexible: Existing systems mostly use obstacle avoidance algorithms with fixed parameters, which cannot be configured in a gradient according to the actual situation of the production line. When facing complex environments, they are prone to obstacle avoidance failure or over-avoidance, which leads to reduced efficiency.

[0004] Lack of visual editing capabilities: The generated robot program cannot be visually edited and optimized, making it difficult for users to adjust the program according to actual needs, resulting in long program debugging cycles and poor adaptability.

[0005] The simulation is out of touch with the real environment: Existing simulation systems cannot accurately simulate the physical process of laser cleaning, the cleaning effect is not realistic, and it cannot effectively guide actual production. The simulation results differ greatly from the real cleaning effect.

[0006] Insufficient virtual-physical linkage capability: There is a lack of effective two-way data communication mechanism between the virtual environment and physical devices, making it impossible to achieve true digital twin and closed-loop control, and the results of virtual debugging are difficult to be directly applied to actual production.

[0007] Low system integration: There is a lack of complete system solutions from model import, path planning, trajectory planning, cleaning simulation, program simulation, program distribution, program execution, data monitoring to data storage. Each link is independent of the others, and data flow is not smooth.

[0008] Poor workpiece adaptability: The system has difficulty adapting quickly to the switching of workpieces of different sizes and shapes, and lacks a universal workpiece compatibility mechanism, resulting in insufficient production line flexibility.

[0009] Lack of process knowledge base: The system cannot accumulate and reuse laser cleaning process knowledge, and process parameters need to be explored again for each new workpiece, which is inefficient.

[0010] Real-time control schemes based on 3D vision have limitations: Some current solutions use 3D vision techniques such as RGBD cameras and 3D point cloud reconstruction to directly guide laser cleaning operations, driving robot movement and cleaning head pose adjustment through real-time sensor data. However, these schemes heavily rely on physical sensing devices (such as depth cameras and laser rangefinders) and mechanical compensation mechanisms (such as sliding rail adjustment devices), resulting in high system hardware costs and complex maintenance. Furthermore, the lack of pre-simulation verification capabilities in virtual environments prevents sufficient simulation verification and optimization of cleaning paths, obstacle avoidance strategies, and process parameters before actual production, leading to long on-site debugging cycles and high trial-and-error costs. In addition, these schemes typically employ a single geometric coverage path planning algorithm (such as the grid method), failing to fully integrate the laser cleaning head's process characteristic parameters (such as optimal working distance, incident angle, and spot energy distribution) for path optimization, making it difficult to achieve optimal cleaning quality.

[0011] Existing solutions lack a closed-loop linkage mechanism between virtual design and physical production: whether it is a traditional teaching programming solution or a real-time control solution based on 3D vision, it is a one-way design → execution or perception → control process. It lacks a closed-loop mechanism to feed back physical operation data to the virtual environment for continuous optimization. It cannot realize the full-process digital twin management from production line planning, simulation trial operation, real machine debugging to production operation, which restricts the system's continuous optimization capability and intelligence level. Summary of the Invention

[0012] The purpose of this application is to provide an automated laser cleaning method, system, and medium based on 3D graphics. By using 3D graphics algorithms and laser cleaning process knowledge, the optimal cleaning path can be automatically generated according to the appearance characteristics of the workpiece, thereby improving cleaning quality and efficiency.

[0013] This application also provides an automated laser cleaning method based on 3D graphics, including: A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or workpieces are scanned to obtain 3D models of the workpieces and configure the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms. Based on the interference obstacle avoidance algorithm, cleaning path and robot kinematic model, a gradient-based obstacle avoidance algorithm is used to generate robot motion trajectory, and the generated robot motion trajectory is then visualized and optimized by the program. Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment; Based on the cleaning simulation execution program, the cleaning process data is collected in real time and the cleaning process data is stored and fed back to the virtual environment.

[0014] Optionally, in the 3D graphics-based automated laser cleaning method described in this application embodiment, a digital twin of the production line is constructed based on a three-dimensional virtual environment, a workpiece model is imported or the workpiece is scanned to obtain a three-dimensional model of the workpiece, and the characteristic parameters of the laser cleaning head are configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

[0015] Optionally, in the 3D graphics-based automated laser cleaning method described in this application embodiment, a cleaning path is generated using a 3D graphics algorithm based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, specifically including: Analyze the surface features of the workpiece's 3D model to identify the areas to be cleaned and obstacle areas; The process parameters of the laser cleaning head are obtained, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position. Based on the optimal working distance range of the laser cleaning head, the optimal distance range between the cleaning head and the workpiece surface is calculated; Based on the workpiece surface normal vector and the incident angle requirement of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated; Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned, thus obtaining the cleaning path.

[0016] Optionally, in the 3D graphics-based automated laser cleaning method described in this application embodiment, a gradient-based obstacle avoidance algorithm is used to generate the robot's motion trajectory based on the interference obstacle avoidance algorithm, the cleaning path, and the robot's kinematic model. Specifically, this includes: Construct a 3D collision model of the production line environment, including the robot body, laser cleaning head, workpiece, positioner, and worktable; Set gradient obstacle avoidance parameters, including safety distance level, obstacle avoidance response speed and obstacle avoidance priority; Based on the cleaning path point sequence, the robot joint angle corresponding to each path point is calculated using the forward and inverse kinematics algorithm. To detect potential collisions during robot movement, a hierarchical bounding box or convex hull algorithm is used for collision detection. When a potential collision is detected, the robot trajectory is generated based on the gradient obstacle avoidance parameters, and the robot trajectory is adjusted using path replanning or attitude adjustment strategies. The adjusted robot trajectory is smoothed to obtain the optimized robot motion trajectory.

[0017] Optionally, in the 3D graphics-based automated laser cleaning method described in this application embodiment, the generated robot motion trajectory is visually edited and the program is optimized, specifically including: The robot's motion trajectory is displayed graphically, including joint angle curves, end effector trajectory, velocity curves, and acceleration curves; The system uses a graphical interface to generate a visual editing interface, allowing users to add, delete, and modify trajectory points. It supports code-level editing of robot programs to obtain trajectory effects. Code-level editing includes motion instructions, IO control instructions, and process parameter setting instructions. Real-time preview of the edited trajectory effect, and collision detection and reachability verification; Perform syntax checks and logic verifications on the edited program, and generate a verification report.

[0018] Optionally, in the 3D graphics-based automated laser cleaning method described in this application embodiment, cleaning simulation is performed in a three-dimensional virtual environment based on the physical principles of the laser cleaning head, specifically including: Establish a virtual asset model of the laser cleaning head, including laser beam shape, energy distribution, effective range, and cleaning effect model; Based on the physical principles of laser cleaning, including the interaction between laser and materials, material removal mechanisms, and changes in surface morphology, a calculation model for cleaning effect is established. During the cleaning simulation, the cleaning effect in the laser beam action area is calculated in real time based on the cleaning effect calculation model, including the material removal depth, surface roughness change, and oxide layer removal degree. Based on the physical rendering algorithm, the surface of the workpiece is dynamically colored and rendered according to the calculation results of the cleaning effect, simulating the visual changes before and after cleaning.

[0019] Secondly, embodiments of this application provide an automated laser cleaning system based on 3D graphics. The system includes a memory and a processor. The memory includes a program for an automated laser cleaning method based on 3D graphics. When the program for the automated laser cleaning method based on 3D graphics is executed by the processor, it performs the following steps: A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or workpieces are scanned to obtain 3D models of the workpieces and configure the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms. Based on the interference obstacle avoidance algorithm, cleaning path and robot kinematic model, a gradient-based obstacle avoidance algorithm is used to generate robot motion trajectory, and the generated robot motion trajectory is then visualized and optimized by the program. Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment; Based on the cleaning simulation execution program, the cleaning process data is collected in real time and the cleaning process data is stored and fed back to the virtual environment.

[0020] Optionally, in the 3D graphics-based automated laser cleaning system described in this application embodiment, a digital twin of the production line is constructed based on a three-dimensional virtual environment, a workpiece model is imported or the workpiece is scanned to obtain a three-dimensional model of the workpiece, and the characteristic parameters of the laser cleaning head are configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

[0021] Optionally, in the 3D graphics-based automated laser cleaning system described in this application embodiment, a cleaning path is generated using a 3D graphics algorithm based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, specifically including: Analyze the surface features of the workpiece's 3D model to identify the areas to be cleaned and obstacle areas; The process parameters of the laser cleaning head are obtained, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position. Based on the optimal working distance range of the laser cleaning head, the optimal distance range between the cleaning head and the workpiece surface is calculated; Based on the workpiece surface normal vector and the incident angle requirement of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated; Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned, thus obtaining the cleaning path.

[0022] Thirdly, embodiments of this application also provide a computer-readable storage medium, which includes a 3D graphics-based automated laser cleaning method program. When the 3D graphics-based automated laser cleaning method program is executed by a processor, it implements the steps of the 3D graphics-based automated laser cleaning method as described in any of the preceding claims.

[0023] As can be seen from the above, the automated laser cleaning method, system, and medium based on 3D graphics provided in this application construct a digital twin of the production line in a three-dimensional virtual environment, import the workpiece model or scan the workpiece to obtain the three-dimensional model of the workpiece, and configure the characteristic parameters of the laser cleaning head; based on the appearance features of the workpiece and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms; based on the interference obstacle avoidance algorithm, the cleaning path, and the robot kinematics model, a gradient-based obstacle avoidance algorithm is used to generate the robot motion trajectory, and the generated robot motion trajectory is visualized and optimized; based on the physical principles of the laser cleaning head, cleaning simulation is performed in a three-dimensional virtual environment; based on the cleaning simulation, the program is executed to simulate and collect cleaning process data in real time, and the cleaning process data is stored and fed back to the virtual environment; through 3D graphics algorithms and laser cleaning process knowledge, the optimal cleaning path can be automatically generated according to the appearance features of the workpiece, thereby improving cleaning quality and efficiency. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart illustrating an automated laser cleaning method based on 3D graphics provided in this application embodiment; Figure 2 A flowchart illustrating the cleaning path generation method for the automated laser cleaning method based on 3D graphics provided in this application embodiment; Figure 3 A flowchart of the gradient obstacle avoidance algorithm for the automated laser cleaning method based on 3D graphics provided in this application embodiment; Figure 4 A flowchart of the color rendering algorithm for the cleaning effect of the automated laser cleaning method based on 3D graphics provided in the embodiments of this application; Figure 5 A flowchart illustrating the virtual-real interconnected data flow control of the automated laser cleaning method based on 3D graphics provided in this application embodiment; Figure 6 A framework diagram of an automated laser cleaning system based on 3D graphics provided for embodiments of this application; Figure 7 A schematic diagram illustrating a typical application scenario of the automated laser cleaning system based on 3D graphics provided in this application embodiment; Figure 8 A schematic diagram of the cleaning path planning for the automated laser cleaning method based on 3D graphics provided in the embodiments of this application; Figure 9 A schematic diagram of gradient obstacle avoidance for an automated laser cleaning method based on 3D graphics provided in an embodiment of this application; Figure 10 This is a rendering comparison of the cleaning effect of the automated laser cleaning method based on 3D graphics provided in the embodiments of this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0027] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0028] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating an automated laser cleaning method based on 3D graphics, as described in some embodiments of this application. This automated laser cleaning method based on 3D graphics is used in a terminal device and includes the following steps: S101: Construct a digital twin of the production line based on a three-dimensional virtual environment, import workpiece models or scan workpieces to obtain three-dimensional workpiece models, and configure the characteristic parameters of the laser cleaning head. S102, based on the workpiece appearance features and the characteristic parameters of the laser cleaning head, uses 3D graphics algorithms to generate the cleaning path; S103, based on the interference obstacle avoidance algorithm, cleaning path and robot kinematics model, uses a gradient-based obstacle avoidance algorithm to generate robot motion trajectory, and performs visual editing and program optimization on the generated robot motion trajectory; S104, based on the physical principles of laser cleaning heads, performs cleaning simulation in a three-dimensional virtual environment; S105 is based on a cleaning simulation program to collect cleaning process data in real time and store the cleaning process data back to the virtual environment.

[0029] This method enables intelligent management of the entire process from production line planning, simulated trial operation, real machine debugging to production operation. It guides real production through virtual design, and feeds back physical operation data to the virtual environment to form a closed-loop control, which promotes continuous performance monitoring and optimization and solves the flexible rust removal and cleaning needs of end users such as automotive parts manufacturers, OEMs, and machining plants.

[0030] Specifically, the execution program simulation includes the joint simulation of equipment such as robot motion, I / O signals, and the working status of the cleaning head.

[0031] Specifically, this includes: robot motion simulation: Based on the robot's kinematic model, calculate the angle, velocity, and acceleration of each joint at different time points; The robot's motion status is displayed in real time in a 3D virtual environment, including the position, attitude, and trajectory of each link; Detects problems such as collisions, exceeding limits, and singular configurations during robot movement.

[0032] IO signal simulation: Simulate various I / O signals, including input signals (sensor signals, button signals, etc.) and output signals (control signals, status signals, etc.); Establish the logical relationship between I / O signals and robot movement and cleaning head operation; Simulate the timing of I / O signals to verify the correctness of the program logic.

[0033] Simulation of the working state of the cleaning head: Simulate the on / off state, power adjustment, scanning speed and other parameters of the cleaning head; Calculate the cleaning effect based on the condition of the cleaning head; Simulate cleaning head malfunctions and abnormal situations to verify the system's fault tolerance capability.

[0034] Co-simulation: Synchronize the simulation status of all devices to ensure consistent simulation time; Handles interactions and dependencies between devices; Generate a simulation report, including motion time, cleaning effect, energy consumption, potential problems, etc.

[0035] Furthermore, it also includes distributing the optimized program to terminal devices to achieve two-way data communication, variable communication, SDK direct control, and data transmission between virtual and physical systems.

[0036] Specifically, this includes: establishing communication connections between the virtual environment and physical devices. It supports multiple communication protocols, including TCP / IP, Modbus, EtherCAT, OPC UA, and robot-specific protocols (such as KUKA Ethernet KRL, ABB PC Interface, etc.). Establish a communication connection, monitor the connection status, and handle any anomalies.

[0037] Convert robot programs in a virtual environment into a program format that can be recognized by physical devices: Based on the target robot controller type, convert the general program format to a specific format (such as KUKA KRL, ABRAPID, FANUC TP, etc.). Perform format verification to ensure the converted program has the correct format; Supports segmented distribution and incremental updates of the program.

[0038] Through a variable mapping mechanism, bidirectional binding and real-time synchronization between virtual environment variables and physical device variables are achieved: Establish a variable mapping table to define the correspondence between virtual variables and physical variables; Achieve bidirectional synchronization of variables: update physical device variables when virtual environment variables change, and update virtual environment variables when physical device variables change. It supports data type conversion and unit conversion for variables.

[0039] Provides an SDK interface to support external systems in directly controlling virtual environments and physical devices: It provides API interfaces to support operations such as program calls, parameter settings, status queries, and data reading; Supports multiple programming languages ​​(such as C++, C#, Python, Java, etc.); Provide API documentation and sample code.

[0040] Real-time data transmission: Transmits motion data such as robot position, speed, and acceleration; Transmits device data such as I / O status and sensor data; Transmit process data such as cleaning parameters and cleaning results; Enables real-time data transmission and storage.

[0041] According to an embodiment of the present invention, a digital twin of the production line is constructed based on a three-dimensional virtual environment, a workpiece model is imported or the workpiece is scanned to obtain a three-dimensional model of the workpiece, and the characteristic parameters of the laser cleaning head are configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

[0042] It should be noted that by generating the cleaning path based on 3D graphics algorithms and laser cleaning process knowledge, the optimal working range and incident angle of the laser cleaning head can be fully considered, thereby improving cleaning quality and efficiency.

[0043] like Figure 2 As shown in the embodiment of the present invention, based on the workpiece appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using a 3D graphics algorithm, specifically including: S201 analyzes the surface features of the workpiece's 3D model, extracts surface normals, curvature, and uneven areas, and identifies the areas to be cleaned and obstacle areas.

[0044] S202, Obtain the process parameters of the laser cleaning head, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position.

[0045] S203 calculates the optimal distance between the cleaning head and the workpiece surface based on the optimal working distance range of the laser cleaning head, ensuring that the distance is within the optimal working range.

[0046] S204, based on the workpiece surface normal vector and the incident angle requirements of the laser cleaning head, calculates the optimal attitude angle of the cleaning head to ensure that the incident angle meets the process requirements.

[0047] S205. Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned.

[0048] S206 optimizes the cleaning path to ensure it is continuous and smooth, and meets the process constraints of the laser cleaning head.

[0049] It should be noted that by generating the cleaning path based on 3D graphics algorithms and laser cleaning process knowledge, the optimal working range and incident angle of the laser cleaning head can be fully considered, thereby improving cleaning quality and efficiency.

[0050] It should be noted that the surface features of the workpiece's 3D model are analyzed as follows: Extract the geometric information of the workpiece surface, including surface normal vector, curvature, and concave and convex regions; Identify areas to be cleaned and obstacle areas, either through user annotation or automatic identification (e.g., based on color, texture, or geometric features). Analyze surface complexity, including different types of regions such as planes, curved surfaces, and complex curved surfaces.

[0051] Based on the optimal working distance range of the laser cleaning head, the optimal distance between the cleaning head and the workpiece surface is calculated: For each point P_i on the workpiece surface, with normal vector n_i and cleaning head position H, the distance from the cleaning head to the surface is: d_i = |(P_i - H)·n_i| The optimal distance should satisfy: d_min ≤ d_i ≤ d_max Based on the workpiece surface normal vector and the incident angle requirements of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated: If the incident direction vector of the cleaning head is l, and the surface normal vector is n, then the incident angle is: θ = arccos(|l·n|) The optimal orientation should satisfy: θ_min ≤ θ ≤ θ_max, and θ is usually required to be close to 90° (vertical incidence) to obtain the best cleaning effect.

[0052] Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a cleaning path: An equal-spacing scanning strategy is adopted: the path point density is calculated based on the cleaning head spot diameter d and the scanning spacing s to ensure coverage of all areas to be cleaned without omissions or overlaps; For planar areas, a straight scanning path is used; For curved regions, adaptive path planning is used to adjust the path density and direction according to the curvature; For complex areas, layered scanning or multi-angle scanning strategies are adopted.

[0053] Path planning algorithm flow: Discretize the region to be cleaned into a grid point set G = {g_1,g_2, ..., g_n}; Based on the coverage area of ​​the cleaning head, calculate whether each grid point has been covered; Use a greedy algorithm or a genetic algorithm to generate the shortest path covering all grid points; The path is smoothed to ensure it is continuous and without abrupt changes.

[0054] Optimize the cleaning path: Path continuity check: Ensures smooth connections between adjacent path points; Process constraint verification: Verify whether each path point meets the process constraints (distance, angle, etc.) of the laser cleaning head. Path length optimization: The TSP (Traveling Salesman Problem) algorithm is used to optimize the path order and reduce wasted time.

[0055] like Figure 3 As shown in the embodiment of the present invention, based on the interference obstacle avoidance algorithm, the cleaning path, and the robot kinematics model, a gradient-based obstacle avoidance algorithm is used to generate the robot's motion trajectory, specifically including: S301: Construct a 3D collision model of the production line environment, establish colliders for all objects that may interfere, and represent the colliders using hierarchical bounding box or convex hull algorithms.

[0056] S302 allows you to set gradient obstacle avoidance parameters, including safety distance level, obstacle avoidance response speed, and obstacle avoidance priority, and supports production line-level custom configuration.

[0057] S303 calculates the robot joint angle corresponding to each path point based on the cleaning path point sequence using forward and inverse kinematics algorithms.

[0058] S304 detects potential collisions during robot movement and uses a hierarchical collision detection algorithm to quickly determine whether there is a collision risk.

[0059] S305: When a potential collision is detected, the robot trajectory is adjusted according to the gradient obstacle avoidance parameters, and a path replanning or attitude adjustment strategy is adopted to avoid the collision.

[0060] S306 smooths the adjusted trajectory to ensure continuous and stable robot motion, meeting kinematic constraints and dynamic limitations.

[0061] It should be noted that the gradient obstacle avoidance algorithm supports production line-level custom configuration. Different levels of obstacle avoidance parameters can be set according to different scenarios to achieve flexible and efficient obstacle avoidance strategies and improve system adaptability and safety.

[0062] Specifically, this includes: constructing a 3D collision model of the production line environment. Establish collision bodies for all objects that may interfere, including the links of the robot body, the laser cleaning head, the workpiece, the positioner, the worktable, and surrounding equipment; Using hierarchical bounding boxes (AABB, OBB) or convex hull algorithms to represent colliders improves collision detection efficiency; Establish a hierarchical structure for colliders to support fast collision lookup.

[0063] Configure gradient obstacle avoidance parameters, supporting custom configuration at the production line level: Safety distance levels: Multiple safety distance levels can be set (e.g., L1: 5mm, L2: 10mm, L3: 20mm), with different levels corresponding to different obstacle avoidance strategies; Obstacle avoidance response speed: Different response speeds are set according to the collision risk level, with a fast response in high-risk areas and a smooth transition in low-risk areas; Obstacle avoidance priority: Set the obstacle avoidance priority for different objects, such as workpiece > positioner > worktable > surrounding equipment; Obstacle avoidance strategies include path replanning, attitude adjustment, speed reduction, and pause / wait.

[0064] Based on the sequence of clean path points, the robot joint angles corresponding to each path point are calculated using forward and inverse kinematics algorithms. For a six-axis robot, the kinematic model is established using the DH parameter method: Forward kinematics: Calculate the end-effector pose T_6^0 based on joint angles θ = [θ_1, θ_2, ..., θ_6]. T_6^0 = T_1^0(θ_1)·T_2^1(θ_2) · ... · T_6^5(θ_6) Inverse kinematics: Calculate the joint angle θ based on the end pose T_6^0, and solve it using analytical or numerical iteration methods.

[0065] Detecting potential collisions during robot movement: For each path point, calculate the position and orientation of each link of the robot; A hierarchical collision detection algorithm is used to quickly determine whether there is a collision risk; Calculate the collision distance and determine the collision risk level.

[0066] When a potential collision is detected, adjust the robot's trajectory according to the gradient-based obstacle avoidance parameters: If the collision distance < L1 (high risk), adopt a path replanning strategy and recalculate the obstacle avoidance path; If L1 ≤ collision distance < L2 (medium risk), adopt an attitude adjustment strategy to fine-tune the robot's attitude to avoid obstacles; If L2 ≤ collision distance < L3 (low risk), adopt a speed reduction strategy to reduce the movement speed to improve safety; If the collision distance ≥ L3 (safe), keep the original trajectory unchanged.

[0067] Obstacle avoidance path replanning algorithm: Identify the collision area and obstacles; Search for feasible path points around the collision area; Adopt the A* algorithm or RRT algorithm to plan the obstacle avoidance path; Smooth the obstacle avoidance path.

[0068] Smooth the adjusted trajectory: Adopt spline interpolation algorithms (such as B-spline, Catmull-Rom spline) to interpolate the trajectory points to ensure the continuity and smoothness of the trajectory; Verify whether the trajectory meets the robot's kinematic constraints (joint angle limits, joint speed limits, joint acceleration limits); Verify whether the trajectory meets the dynamic constraints (joint torque limits); Optimize the trajectory to reduce the movement time and energy consumption.

[0069] According to the embodiments of the present invention, perform visual editing and program optimization on the generated robot motion trajectory, specifically including: Display the robot motion trajectory in a graphical manner, including joint angle curves, end trajectories, speed curves, and acceleration curves; Based on the graphical display, generate a visual editing interface that supports users to add, delete, and modify trajectory points; Support code-level editing of the robot program to obtain the trajectory effect. The code-level editing includes motion instructions, IO control instructions, and process parameter setting instructions; Preview the edited trajectory effect in real time and perform collision detection and reachability verification; Perform syntax checking and logical verification on the edited program and generate a verification report.

[0070] Specifically include: Display the robot motion trajectory in a graphical manner: Display the robot motion trajectory in a three-dimensional virtual environment, including joint angle curves, end trajectories, speed curves, and acceleration curves; Provides a timeline view to show the status of each joint at different points in time; Provides a list of trajectory points, displaying detailed information for each trajectory point (position, attitude, velocity, acceleration, etc.).

[0071] Provides a visual editing interface: It supports adding, deleting, and modifying trajectory points, and users can directly drag and drop trajectory points in the 3D view to adjust them; Supports insertion, deletion, and replacement operations for trajectory segments; Provides undo / redo functionality and supports editing and rewinding history.

[0072] Supports code-level editing of robot programs: Convert the trajectory into program code that the robot controller can recognize (such as KUKA KRL, ABB RAPID, FANUCTP, etc.). Provides a code editor that supports syntax highlighting, auto-completion, and error checking; It supports adding motion commands (such as MoveL, MoveJ, MoveC, etc.), IO control commands (such as Set, Reset, Wait, etc.), and process parameter setting commands.

[0073] Real-time preview of the edited trajectory effect: The 3D view is updated in real time during the editing process to display the edited trajectory effect; Collision detection and accessibility verification are performed to ensure the safety and feasibility of the edited trajectory. Display warnings and error messages to alert the user to potential problems.

[0074] Perform syntax checks and logic verifications on the edited program: Check if the program syntax is correct and conforms to the syntax specifications of the robot controller; Check whether the program logic is reasonable, such as the sequence of motion instructions and the timing of I / O signals; Generate a program verification report, listing the problems found and recommendations.

[0075] According to an embodiment of the present invention, based on the physical principles of a laser cleaning head, a cleaning simulation is performed in a three-dimensional virtual environment, specifically including: Establish a virtual asset model of the laser cleaning head, including laser beam shape, energy distribution, effective range, and cleaning effect model; Based on the physical principles of laser cleaning, including the interaction between laser and materials, material removal mechanisms, and changes in surface morphology, a calculation model for cleaning effect is established. During the cleaning simulation, the cleaning effect in the laser beam action area is calculated in real time based on the cleaning effect calculation model, including the material removal depth, surface roughness change, and oxide layer removal degree. Based on the physical rendering algorithm, the surface of the workpiece is dynamically colored and rendered according to the calculation results of the cleaning effect, simulating the visual changes before and after cleaning.

[0076] like Figure 4 As shown, it should be noted that the color rendering algorithm for simulating the cleaning effect in a 3D virtual environment, based on the physical principles of the laser cleaning head, specifically includes: S401, Establish a virtual asset model of the laser cleaning head, including laser beam shape, energy distribution, effective range, and cleaning effect model.

[0077] S402, based on the physical principles of laser cleaning, including the interaction between laser and materials, material removal mechanisms, and changes in surface morphology, establishes a calculation model for cleaning effect.

[0078] S403 calculates the cleaning effect in the laser beam action area in real time during the cleaning simulation process, including the material removal depth, surface roughness change, and oxide layer removal degree.

[0079] S404 employs a physically based rendering (PBR) algorithm to dynamically color and render the workpiece surface based on the cleaning effect calculation results, simulating the visual changes before and after cleaning.

[0080] The S405 supports multiple rendering modes, including real-time preview mode and high-quality rendering mode, to meet the needs of different application scenarios.

[0081] A virtual asset model is established based on the physical principles of laser cleaning to achieve realistic cleaning effect simulation and color rendering, effectively guiding actual production.

[0082] Specifically, this includes: establishing a virtual asset model for the laser cleaning head. Laser beam geometry model: Establish the three-dimensional geometry of the laser beam, including parameters such as beam diameter, divergence angle, and focal point position; Energy distribution model: Establish a spatial distribution model of laser energy, typically using a Gaussian distribution. I(r) = I_0·exp(-2r² / w²) Where I_0 is the peak power density, r is the distance from the beam center, and w is the beam radius.

[0083] Effective range model: Defines the effective range of the laser beam, including the depth of effect and the area of ​​effect.

[0084] Based on the physical principles of laser cleaning, a calculation model for cleaning effectiveness is established: The physical processes of laser cleaning include: Laser-material interactions include photothermal effects, photochemical effects, and photomechanical effects. Material removal mechanisms include ablation, evaporation, and peeling. Surface morphology changes: including changes in roughness, removal of oxide layer, and improvement in surface cleanliness.

[0085] Cleaning effect calculation model: Material removal depth: calculated based on laser energy density and interaction time. h = f(I, t, α, ρ) Where I is the laser energy density, t is the interaction time, α is the material absorption coefficient, and ρ is the material density.

[0086] Cleaning efficiency: calculated based on cleaning area and cleaning time. η = A_cleaned / (v·t) Where A_cleaned is the cleaned area, v is the scanning speed, and t is the cleaning time.

[0087] During the cleaning simulation, the cleaning effect in the laser beam's action area is calculated in real time: The workpiece surface is discretized into a grid, and each grid point records the cleaning status (not cleaned, cleaning in progress, cleaned) and cleaning parameters (removal depth, roughness, cleanliness, etc.). The cleaning effect of each grid point within the effective area is calculated based on the position and energy distribution of the laser beam. Update the cleaning status and cleaning parameters of the grid points.

[0088] A physically based rendering (PBR) algorithm is used to dynamically color and render the workpiece surface based on the calculation results of the cleaning effect. Rendering algorithm flow: Calculate the cleaning effect parameters (removal depth, roughness, cleanliness, etc.) for each grid point; Calculate surface material properties (diffuse reflection coefficient, specular reflection coefficient, roughness, etc.) based on cleaning effect parameters; Rendering is performed using the PBR rendering pipeline: Calculate lighting (ambient light, direct light, indirect light); Calculate material response (diffuse reflection, specular reflection); By mixing the material properties before and after cleaning, a smooth transition can be achieved; The rendering results are updated in real time, displaying the dynamic effects of the cleaning process.

[0089] Material property calculation formula: Material before cleaning: M_before = {k_d_before, k_s_before, roughness_before} Material after cleaning: M_after = {k_d_after, k_s_after, roughness_after} Hybrid material: M = (1 - α) · M_before + α · M_after Where α represents the cleaning completion rate (between 0 and 1).

[0090] The specific parameters are explained below: 1. Material Collection: M_before: The set of material properties before cleaning, representing the surface optical properties of the workpiece before laser cleaning (such as the state of rust or dirt). M_after: The set of material properties after cleaning, representing the surface optical properties of the workpiece after it has been completely cleaned (such as the state of exposed metallic luster). M: The current set of mixed material properties, representing the real-time visual material properties at a certain point during the cleaning process.

[0091] 2. Physical parameters (corresponding to core attributes in the PBR rendering workflow): k_d (Diffuse Coefficient): Diffuse reflection coefficient. k_d_before is the diffuse reflection intensity before cleaning (e.g., dark red rust), and k_d_after is the diffuse reflection intensity after cleaning (e.g., the original color of the metal). k_s (Specular Coefficient): Specular reflection coefficient. k_s_before is the coefficient before cleaning (weak reflection), and k_s_after is the coefficient after cleaning (strong reflection, such as a mirror). Roughness: Surface roughness. Roughness_before is the roughness before cleaning (highlight dispersion), and roughness_after is the roughness after cleaning (highlight convergence).

[0092] 3. Mixed logic: The formula uses linear interpolation (Lerp) logic. As the laser scans, α gradually changes from 0 to 1, and the visual effect smoothly transitions from rust to metallic luster.

[0093] Supports multiple rendering modes: Real-time preview mode: Employs a simplified rendering algorithm to ensure real-time performance, used for interactive previewing; High-quality rendering mode: Employs the full PBR rendering algorithm to generate high-quality rendered images for result display and report generation.

[0094] Please refer to Figure 5 , Figure 5 This is a flowchart illustrating the virtual-real interconnection data flow control of an automated laser cleaning method based on 3D graphics, according to one embodiment of this application. According to this embodiment, the optimized program is sent to the terminal device to achieve bidirectional data communication between the virtual and real interconnection, specifically including: S501 establishes communication connections between virtual environments and physical devices, supporting multiple communication protocols, including TCP / IP, Modbus, EtherCAT, and OPC UA.

[0095] S502 converts robot programs in a virtual environment into a program format that can be recognized by physical devices, including robot controller programs and PLC programs.

[0096] S503 achieves bidirectional binding and real-time synchronization between virtual environment variables and physical device variables through a variable mapping mechanism.

[0097] The S504 provides an SDK interface that allows external systems to directly control virtual environments and physical devices.

[0098] The S505 transmits data in real time, including robot position, speed, IO status, cleaning parameters, and sensor data, enabling bidirectional data exchange between the virtual environment and physical devices.

[0099] It should be noted that this enables two-way data communication between the virtual environment and physical devices, supports variable communication, direct SDK control, and data transfer, forming a true digital twin and closed-loop control.

[0100] According to an embodiment of the present invention, please refer to Figure 6 , Figure 6 This is a system architecture diagram of an automated laser cleaning system based on 3D graphics in some embodiments of this application.

[0101] The system adopts a modular design and includes the following core modules: Model Editor Module: Supports CAD model import, 3D scan model processing, model repair and optimization.

[0102] Simulation module setup: Provides functions for production line layout design, equipment configuration, and environment setup.

[0103] Virtual debugging module: Enables program simulation, collision detection, trajectory optimization, and process verification.

[0104] Virtual-physical linkage module: Enables two-way communication, data synchronization, program distribution, and real-time control between the virtual environment and physical devices.

[0105] Data management module: Enables data collection, storage, analysis, and feedback.

[0106] Data management between modules is achieved through an ECS architecture. Entities are uniquely identified by GUIDs, data is stored through Components, logic is processed through Systems, and external interfaces are provided through RunContext.

[0107] Please refer to Figure 7 , Figure 7 This is a schematic diagram of a typical application scenario of an automated laser cleaning method based on 3D graphics in some embodiments of this application.

[0108] A typical application scenario involves adding a laser cleaning head to a six-axis robot, configuring a positioner on the worktable, and placing the workpiece at the center of the positioner. The six-axis robot is located on one side of the worktable, and its working radius covers the entire work area. The laser cleaning head is installed at the end of the robot and can be quickly replaced using a quick-change device. The positioner is located in the center of the worktable, supports dual-axis motion, and can adjust the workpiece posture; The workpiece is fixed on the positioner, and the angle and position of the workpiece relative to the cleaning head can be adjusted by moving the positioner.

[0109] The system can automatically plan the coordinated movement of the robot and the positioner to achieve all-round cleaning of the workpiece.

[0110] Please refer to Figure 8 , Figure 8 This is a schematic diagram of the cleaning path planning of an automated laser cleaning method based on 3D graphics in some embodiments of this application.

[0111] The cleaning path planning diagram illustrates the distribution of cleaning path points generated based on workpiece surface features: The workpiece surface is discretized into grid points, and each grid point records surface information (normal vector, curvature, etc.). Based on the coverage area of ​​the cleaning head, generate a path point sequence that covers all areas to be cleaned; The path points are connected in the optimal order to form a continuous cleaning path; For complex areas, layered scanning or multi-angle scanning strategies are adopted.

[0112] Please refer to Figure 9 , Figure 9 This is a schematic diagram of gradient obstacle avoidance for an automated laser cleaning method based on 3D graphics in some embodiments of this application.

[0113] The gradient obstacle avoidance diagram illustrates the obstacle avoidance strategies corresponding to different safety distance levels: L1 area (high risk, distance <5mm): Use path replanning strategy to completely avoid obstacles; L2 region (medium risk, 5mm≤distance<10mm): Use attitude adjustment strategy to fine-tune robot attitude; L3 area (low risk, 10mm≤distance<20mm): adopt a speed reduction strategy to reduce movement speed; Safe zone (distance ≥ 20mm): Maintain the original trajectory.

[0114] Please refer to Figure 10 , Figure 10 These are rendering comparisons of the cleaning effects of an automated laser cleaning method based on 3D graphics, as described in some embodiments of this application.

[0115] The before-and-after renderings of the cleaning process show a visual comparison of the effects before and after cleaning: Before cleaning: The workpiece surface has an oxide layer (such as rust, stains, etc.), is rough, and has low reflectivity; During cleaning: The cleaning area gradually becomes brighter, the surface becomes smoother, and the reflectivity gradually increases; After cleaning: The workpiece surface has a metallic color, is smooth, and has high reflectivity.

[0116] Physically based rendering algorithms can realistically simulate the visual changes in the cleaning effect.

[0117] As can be seen from the above, the automated laser cleaning method, system, and medium based on 3D graphics provided in this application construct a digital twin of the production line based on a three-dimensional virtual environment, import the workpiece model, and configure the laser cleaning head parameters; generate the cleaning path based on 3D graphics algorithms; generate the robot motion trajectory using a gradient obstacle avoidance algorithm; perform visual editing and program optimization; achieve color rendering of the cleaning effect; execute program simulation; realize virtual-real joint control; and provide real-time monitoring and data feedback. Through virtual design, it can guide real production, and the data of physical operation can also be fed back to the virtual environment to form a closed-loop control, which promotes continuous performance monitoring and optimization, enhances the timeliness and accuracy of decision support, and solves the flexible rust removal and cleaning needs of end users.

[0118] Secondly, embodiments of this application provide an automated laser cleaning system based on 3D graphics. The system includes a memory and a processor. The memory includes a program for an automated laser cleaning method based on 3D graphics. When the program for the automated laser cleaning method based on 3D graphics is executed by the processor, it performs the following steps: A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or workpieces are scanned to obtain 3D models of the workpieces and configure the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms. Based on the interference obstacle avoidance algorithm, cleaning path and robot kinematic model, a gradient-based obstacle avoidance algorithm is used to generate robot motion trajectory, and the generated robot motion trajectory is then visualized and optimized by the program. Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment; Based on the cleaning simulation execution program, the cleaning process data is collected in real time and the cleaning process data is stored and fed back to the virtual environment.

[0119] According to embodiments of the present invention, real-time monitoring of cleaning process data, including cleaning quality, equipment status, and process parameters, and feeding the data back to a virtual environment to form a closed-loop control, specifically including: Real-time monitoring of cleaning process data: Cleaning quality data includes cleaning completion rate, surface roughness, cleanliness, and degree of oxide layer removal. Equipment status data: including robot position, speed, load, temperature, I / O status, cleaning head power, scanning speed, etc.; Process parameter data: including laser power, scanning speed, scanning spacing, number of cleaning cycles, etc.

[0120] Data acquisition and storage: Establish a data acquisition system to collect various data in real time; Store data in a database to support historical data querying and analysis; Create data indexes to improve query efficiency.

[0121] Data analysis and feedback: Analyze the collected data to identify anomalies and potential problems; The analysis results are fed back to the virtual environment to update the status of the digital twin; The cleaning path and process parameters are optimized based on feedback data to form a closed-loop control.

[0122] Generate monitoring report: Generate real-time monitoring reports, displaying the current cleaning status and key indicators; Generate historical data analysis reports, including trend analysis and statistical analysis; Generate anomaly reports to record abnormal situations and handling measures.

[0123] System architecture and module design: The entire system adopts a modular design, including the following core modules: Model Editor Module: Supports importing CAD models, including multiple formats such as STEP, IGES, STL, and OBJ; Supports 3D scanning point cloud processing and mesh reconstruction; Provides functions such as model repair, optimization, and simplification; Supports model annotation and region division.

[0124] Setting up the simulation module: Provides production line layout design tools, supporting drag-and-drop equipment configuration; Supports configuration of equipment parameters, including robot parameters, cleaning head parameters, positioner parameters, etc. It provides environment setup capabilities, including workbenches, safety fences, sensors, etc.

[0125] Virtual debugging module: Implement program simulation functions, including robot motion simulation, I / O simulation, and cleaning head simulation; Provides collision detection and trajectory optimization functions; It supports process validation, including cleaning effect prediction, time estimation, and energy consumption analysis.

[0126] Virtual-to-real control module: Enables two-way communication between the virtual environment and physical devices; Supports multiple communication protocols and interfaces; Provides variable mapping and data synchronization functions; Supports SDK interface and integration with external systems.

[0127] Data flow control methods: The system uses an ECS (Entity-Component-System) architecture for data management. Entity: Each entity (such as robot, cleaning head, workpiece, etc.) is uniquely identified using a GUID.

[0128] Component: Stores data about an entity, such as its location, speed, and state.

[0129] System: Processes data in components and implements business logic.

[0130] RunContext: Provides an interface to the outside world, allowing access to the entity's real-time context information via GUID.

[0131] Advantages of this architecture: Data access is efficient, avoiding redundant data transmission and copying; Decoupling between modules improves the system's flexibility and maintainability; It supports real-time data updates and synchronization.

[0132] Typical application scenario example: Taking a typical scenario as an example: A six-axis robot is equipped with a laser cleaning head, a positioner is configured on the worktable, and the workpiece is placed in the center of the positioner. Production line configuration: Six-axis robot: working radius 2000mm, load capacity 20kg; Laser cleaning head: 1000W power, 2mm spot diameter, optimal working distance 50-100mm; Positioner: Dual-axis positioner, supporting 360° rotation and ±90° tilting; Workbench: Dimensions 2000mm × 1500mm, height 800mm.

[0133] Workpiece processing: Import the workpiece CAD model or scan to obtain a 3D model; The system automatically identifies the area to be cleaned; The cleaning path and robot trajectory are automatically adjusted according to the size and shape of the workpiece.

[0134] Path planning: Cleaning paths are generated based on the workpiece's appearance features. Consider the optimal working distance and incident angle of the laser cleaning head; Optimize the path to ensure coverage of all areas to be cleaned.

[0135] Trajectory generation: Robot trajectory is generated based on the cleaning path and robot kinematics model; A gradient-based obstacle avoidance algorithm is used to avoid collisions. Considering the movement of the positioner, achieve coordinated movement between the robot and the positioner.

[0136] Program Editor: Visualize and edit robot trajectories and programs; Optimize the program to improve efficiency and security.

[0137] Cleaning simulation: Simulate the cleaning process in a virtual environment; It displays the cleaning effect in real time and predicts the cleaning quality.

[0138] Program distribution: The optimized program is then sent to the robot controller and the cleaning head controller. Establish a virtual-real linkage to achieve two-way data exchange.

[0139] Production Operations: Real-time monitoring of the cleaning process; Collect data and feed it back to the virtual environment; Optimize process parameters based on feedback data.

[0140] Workpiece compatibility: The system supports switching between workpieces of different sizes and shapes: Quick switching: By importing a new workpiece model, the system automatically replans the path and trajectory, enabling quick switching.

[0141] Parameter self-adaptation: Automatically adjusts cleaning parameters such as scanning speed and scanning spacing according to the workpiece size.

[0142] Accessibility verification: Automatically verifies whether the robot can reach all cleaning locations, and provides warnings and suggestions for unreachable locations.

[0143] According to an embodiment of the present invention, a digital twin of the production line is constructed based on a three-dimensional virtual environment, a workpiece model is imported or the workpiece is scanned to obtain a three-dimensional model of the workpiece, and the characteristic parameters of the laser cleaning head are configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

[0144] According to an embodiment of the present invention, a cleaning path is generated using a 3D graphics algorithm based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, specifically including: Analyze the surface features of the workpiece's 3D model to identify the areas to be cleaned and obstacle areas; Obtain the process parameters of the laser cleaning head, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position; Based on the optimal working distance range of the laser cleaning head, the optimal distance range between the cleaning head and the workpiece surface is calculated; Based on the workpiece surface normal vector and the incident angle requirement of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated; Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned, thus obtaining the cleaning path.

[0145] A third aspect of the present invention provides a computer-readable storage medium including a 3D graphics-based automated laser cleaning method program, wherein when the 3D graphics-based automated laser cleaning method program is executed by a processor, it implements the steps of the 3D graphics-based automated laser cleaning method as described in any of the above claims.

[0146] This invention discloses an automated laser cleaning method, system, and medium based on 3D graphics. It constructs a digital twin of the production line in a 3D virtual environment, imports a workpiece model or scans the workpiece to obtain its 3D model, and configures the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the laser cleaning head's characteristic parameters, a cleaning path is generated using 3D graphics algorithms. Based on an interference obstacle avoidance algorithm, the cleaning path, and a robot kinematics model, a gradient-based obstacle avoidance algorithm is used to generate the robot's motion trajectory, which is then visualized and optimized. Cleaning simulation is performed in the 3D virtual environment based on the physical principles of the laser cleaning head. The simulation execution program collects cleaning process data in real time and stores the cleaning process data back to the virtual environment. Through 3D graphics algorithms and laser cleaning process knowledge, the optimal cleaning path can be automatically generated based on the workpiece's appearance features, improving cleaning quality and efficiency.

[0147] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0148] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0149] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0150] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0151] Alternatively, if the integrated units of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. An automated laser cleaning method based on 3D graphics, characterized in that, include: A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or workpieces are scanned to obtain 3D models of the workpieces and configure the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms. Based on the interference obstacle avoidance algorithm, cleaning path and robot kinematic model, a gradient-based obstacle avoidance algorithm is used to generate robot motion trajectory, and the generated robot motion trajectory is then visualized and optimized by the program. Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment; Based on the cleaning simulation execution program, the cleaning process data is collected in real time and the cleaning process data is stored and fed back to the virtual environment.

2. The automated laser cleaning method based on 3D graphics according to claim 1, characterized in that, A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or scanned to obtain a 3D model of the workpiece. The characteristic parameters of the laser cleaning head are then configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

3. The automated laser cleaning method based on 3D graphics according to claim 2, characterized in that, Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms, specifically including: Analyze the surface features of the workpiece's 3D model to identify the areas to be cleaned and obstacle areas; The process parameters of the laser cleaning head are obtained, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position. Based on the optimal working distance range of the laser cleaning head, the optimal distance range between the cleaning head and the workpiece surface is calculated; Based on the workpiece surface normal vector and the incident angle requirement of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated; Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned, thus obtaining the cleaning path.

4. The automated laser cleaning method based on 3D graphics according to claim 3, characterized in that, Based on the interference obstacle avoidance algorithm, cleaning path, and robot kinematics model, a gradient-based obstacle avoidance algorithm is used to generate the robot's motion trajectory, specifically including: Construct a 3D collision model of the production line environment, including the robot body, laser cleaning head, workpiece, positioner, and worktable; Set gradient obstacle avoidance parameters, including safety distance level, obstacle avoidance response speed and obstacle avoidance priority; Based on the cleaning path point sequence, the robot joint angle corresponding to each path point is calculated using the forward and inverse kinematics algorithm. To detect potential collisions during robot movement, a hierarchical bounding box or convex hull algorithm is used for collision detection. When a potential collision is detected, the robot trajectory is generated based on the gradient obstacle avoidance parameters, and the robot trajectory is adjusted using path replanning or attitude adjustment strategies. The adjusted robot trajectory is smoothed to obtain the optimized robot motion trajectory.

5. The automated laser cleaning method based on 3D graphics according to claim 4, characterized in that, The generated robot motion trajectory is then visualized, edited, and optimized through a program, specifically including: The robot's motion trajectory is displayed graphically, including joint angle curves, end effector trajectory, velocity curves, and acceleration curves; The system uses a graphical interface to generate a visual editing interface, allowing users to add, delete, and modify trajectory points. It supports code-level editing of robot programs to obtain trajectory effects. Code-level editing includes motion instructions, IO control instructions, and process parameter setting instructions. Real-time preview of the edited trajectory effect, and collision detection and reachability verification; Perform syntax checks and logic verifications on the edited program, and generate a verification report.

6. The automated laser cleaning method based on 3D graphics according to claim 5, characterized in that, Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment, specifically including: Establish a virtual asset model of the laser cleaning head, including laser beam shape, energy distribution, effective range, and cleaning effect model; Based on the physical principles of laser cleaning, including the interaction between laser and materials, material removal mechanisms, and changes in surface morphology, a calculation model for cleaning effect is established. During the cleaning simulation, the cleaning effect in the laser beam action area is calculated in real time based on the cleaning effect calculation model, including the material removal depth, surface roughness change, and oxide layer removal degree. Based on the physical rendering algorithm, the surface of the workpiece is dynamically colored and rendered according to the calculation results of the cleaning effect, simulating the visual changes before and after cleaning.

7. An automated laser cleaning system based on 3D graphics, characterized in that, The system includes a memory and a processor. The memory contains a program for an automated laser cleaning method based on 3D graphics. When the program for the automated laser cleaning method based on 3D graphics is executed by the processor, it performs the following steps: A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or workpieces are scanned to obtain 3D models of the workpieces and configure the characteristic parameters of the laser cleaning head. Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms. Based on the interference obstacle avoidance algorithm, cleaning path and robot kinematic model, a gradient-based obstacle avoidance algorithm is used to generate robot motion trajectory, and the generated robot motion trajectory is then visualized and optimized by the program. Based on the physical principles of laser cleaning heads, cleaning simulation is performed in a three-dimensional virtual environment; Based on the cleaning simulation execution program, the cleaning process data is collected in real time and the cleaning process data is stored and fed back to the virtual environment.

8. The automated laser cleaning system based on 3D graphics according to claim 7, characterized in that, A digital twin of the production line is constructed based on a 3D virtual environment. Workpiece models are imported or scanned to obtain a 3D model of the workpiece. The characteristic parameters of the laser cleaning head are then configured, specifically including: A digital twin of the production line is created based on a 3D virtual environment, including virtual models of all equipment such as robots, laser cleaning heads, positioners, and workbenches, generating a complete virtual production line environment; Import the workpiece model, which supports multiple CAD formats, or obtain the point cloud data of the workpiece through a 3D scanning device, perform point cloud processing and mesh reconstruction, and generate a 3D model of the workpiece. Configure the characteristic parameters of the laser cleaning head, including: laser parameters, optical parameters, process parameters, and working range; Establish a virtual asset model of the laser cleaning head, including the geometry of the laser beam, energy distribution model, and effective range model.

9. The automated laser cleaning system based on 3D graphics according to claim 8, characterized in that, Based on the workpiece's appearance features and the characteristic parameters of the laser cleaning head, a cleaning path is generated using 3D graphics algorithms, specifically including: Analyze the surface features of the workpiece's 3D model to identify the areas to be cleaned and obstacle areas; The process parameters of the laser cleaning head are obtained, including laser power, scanning speed, spot diameter, optimal working distance range, incident angle range, and focal position. Based on the optimal working distance range of the laser cleaning head, the optimal distance range between the cleaning head and the workpiece surface is calculated; Based on the workpiece surface normal vector and the incident angle requirement of the laser cleaning head, the optimal attitude angle of the cleaning head is calculated; Based on the workpiece surface curvature and the coverage area of ​​the cleaning head, a path planning algorithm is used to generate a sequence of cleaning path points covering all areas to be cleaned, thus obtaining the cleaning path.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a 3D graphics-based automated laser cleaning method program, which, when executed by a processor, implements the steps of the 3D graphics-based automated laser cleaning method as described in any one of claims 1 to 6.

Images