Research On Spray Trajectory Planning Of Automobile Panel Based On Machine Vision

Spray painting robots are widely used in surface coating operations of products as important automated production equipment both domestically and internationally.With the increasingly mature development of machine vision technology,research on spray painting robots has gradually shifted from 2D to 3D,incorporating a mechanical system that imitates arms and a visual system that imitates eyes,further improving the spraying efficiency and accuracy of spray painting robots,and has become a research hotspot in the field of robotic spraying.Therefore,this article focuses on a self-made three-degree-of-freedom spray painting robot,utilizing a machine vision system to obtain the coordinate information of the target spraying trajectory,and conducting multi-objective trajectory optimization research based on this.The main research contents are as follows:(1)Structural design and kinematic analysis of the three-degree-of-freedom spray painting robot system.In response to the actual spraying needs of automobile fenders,the design of the three-degree-of-freedom spray painting robot and machine vision system is completed,and the physical structure of the robot is designed in CATIA to determine its size parameters and component composition.Corresponding requirements for mechanical performance indicators and working environment indicators are proposed to ensure operational safety.The kinematic model of the three-degree-of-freedom spray painting robot is established using the D-H parameter method,and kinematic analysis is completed using numerical methods to lay the theoretical foundation for subsequent trajectory planning research.(2)Coordinate positioning of the automobile fender spraying trajectory based on machine vision.In order to obtain accurate position information,camera calibration is completed to realize the conversion between the image coordinate system and the robot coordinate system.The captured color photos of the fender are preprocessed through grayscale conversion,binarization,image denoising,and morphological processing to obtain images with clear features that are easy for the next step of image processing.Then,the Canny edge detection algorithm is used to obtain the edge image of the fender,and the contour extraction algorithm is used to obtain the accurate position information of the fender contour,that is,the coordinate information of the spraying trajectory points.(3)Research on multi-objective trajectory planning algorithm for the three-degree-of-freedom spray painting robot.The "7-5-7" hybrid interpolation algorithm is used for trajectory planning of the joint space manipulator to obtain a continuous and stable spraying trajectory.Based on this trajectory,a multi-constraint objective function is established according to the time-energy-impact performance indicators of the robot during operation,and a hybrid multi-objective NSGA-II(HMONSGA-II)optimization algorithm is proposed to optimize the spraying trajectory for the optimization objectives.A simulation model is built in Matlab/Simulink module to complete the trajectory simulation,and the simulation results show that the HMONSGA-II multi-objective optimization algorithm based on the hybrid interpolation algorithm can achieve the optimal spraying trajectory with shorter time,lower energy consumption,and smaller impact,with obvious optimization performance.(4)Based on the design scheme,the selection of hardware and software is completed,and the experimental platform of the three-degree-of-freedom spray painting robot is built to realize the fusion of machine vision and motion control of the spray painting robot.The optimal time-energy-impact trajectory based on the HMONSGA-II algorithm is tested and analyzed.The experimental results show that the maximum error between the actual trajectory and the simulation trajectory is 8.63%,verifying the effectiveness of the multi-objective trajectory planning algorithm in this article and the correctness of the simulation results.