Research On Feature Recognition And Trajectory Planning For Rough Parts Based On Automatic Programming

Grinding is an important processing method,which largely determines the final quality and surface accuracy of products.At present,the manufacturing industry develops towards intellectualization and automation,so it is the general trend to realize the intellectualization of grinding process.However,at present,the grinding of blanks mainly relies on manual work,which is inefficient and uneven in processing quality.At the same time,the poor production environment endangers the health of workers.At present,some companies have adopted robotic grinding system instead of manual grinding,but the application of these grinding robots is mainly teaching programming.This kind of processing method has poor flexibility and has no ability of visual recognition of workpiece.When replacing the workpiece,the robot needs to be re-teached,which is time-consuming and labor-consuming,and affects the production efficiency.In this paper,based on visual sensing technology,the feature recognition and grinding trajectory planning of the blank parts are studied.On this basis,an automatic programming system of the robot is designed,which achieves the goal of feature recognition,trajectory planning and automatic programming of blank parts,and improves the degree of intelligence and automation of the robot grinding.A visual recognition system for grinding path is established.The camera imaging model is established,and the checkerboard plane calibration method is used to calibrate the camera in the Matlab camera calibration toolbox to get the camera's internal and external parameter matrix.The “Eye-to-Hand” hand-eye relationship is built to solve the “hand-eye relationship”matrix,and the transformation relationship between the image coordinate system and the robot's basic coordinate system is obtained by combining the camera calibration results.The standard workpiece is processed by image processing and edge path information extraction.The standard workpiece image is preprocessed by using MATLAB;Canny algorithm is used for edge detection and image post-processing is used to remove isolated areas.According to the method of equal parameter and equal chord height error,the edge of workpiece is discrete.The coordinates of discrete points in the world coordinate system are obtained by coordinate transformation,and the discrete points are processed by subspline interpolation and fitting.Feature matching and grinding trajectory planning are carried out.The SIFT feature point matching is used to process the blank,and the grinding path planning scheme is studied.A scheme for determining grinding times and grinding paths is proposed.The grinding path trajectory planning is carried out.The kinematics analysis and D-H model of the robot are carried out,and the forward and inverse kinematics of the robot are solved.The trajectory planning simulation of grinding path is carried out,which verified the feasibility of grinding with IRB2400 robot.According to the rules of ABB robot programming instructions and the results of image processing and trajectory planning,an automatic programming platform is designed.The common programming instructions of ABB robot are introduced briefly.The framework of the automatic programming oriented blank feature recognition and trajectory planning system is analyzed.According to the motion instructions of the robot and the method of approximating grinding edge curve by chord segments in this paper,an automatic programming platform is designed.The automatic programming platform generates the robot operation program according to the discrete points of grinding edge,which verifies the validity of the automatic programming platform.