Design And Research Of Robot Laser Punching System Based On Vision Guidance

In the traditional drilling process,it is necessary to use fixtures to position and clamp parts.However,the fixture has restrictions on the hardness of the parts,and some special fixtures cost more.In view of the above problems,this paper combines laser drilling,robot and vision technology,and designs a robot laser drilling control system based on vision guidance.Compared with the traditional punching system,the system designed in this paper does not require fixtures,and the system accuracy meets the actual needs.The specific research contents of this article are as follows:According to the requirements of punching,the hardware required for the punching system is designed.At the same time,referring to the CNC machining process,the entire control system is divided into: communication module,execution module,visual positioning module and trajectory planning module.At the same time in the communication module,according to the communication protocol between the hardware,use Qt-Creator to write a communication system.In the execution module,according to the hardware of the module,the kinematics model of the execution module is established.And according to the complexity of the model,focus on the analysis of the robot's forward and inverse kinematics model.The DH parameter method is used to establish the robot's positive kinematics model.Based on the robot's positive kinematics model,a robot inverse kinematics model based on closed solution,SQP algorithm and BP-SQP algorithm is proposed.And carry out simulation experiments,analyze the accuracy and time complexity of different inverse kinematics models,and determine the use scenarios of different inverse solutions according to their accuracy and time complexity.In the visual positioning module,the visual positioning of the parts is realized based on the imaging model of the camera and the point cloud information of the parts.First,the checkerboard punctuation method is used to determine the internal and external parameter matrix of the camera.Second,use part STL files to obtain part point cloud information.Then,using the external parameter matrix of the camera and the positive kinematics model of the robot,the posture of the camera fixture is calibrated.Finally,according to the posture of the camera fixture,the internal parameter matrix of the camera and the contour of the part,the posture calibration of the part is realized.Simulation experiments are carried out to analyze the accuracy of the vision-guided positioning model.The experimental results show that the model has good accuracy and practicality.In the trajectory planning module,a variable thickness 3D printing algorithm is used to plan the machining trajectory of a single hole,and a genetic algorithm is used to plan the machining trajectory between the punching systems.In the trajectory planning process of the punching system,according to whether the assistance of the motion slide platform is needed,the model is divided into: a small area TSP model that does notrequire the motion slide assist,and a large area TSP model that requires the motion slide assist.In the small-area model,the minimum variation of the total joint radian of the robot is taken as the objective function;in the large-area model,the minimum total movement of the motion slide is the objective function.Finally,according to the flow of each module,a complete control system is designed,and the interface of the control system is written.Taking the car fender punching as the experimental object,the simulation experiment was carried out.The experimental results show that the accuracy and efficiency of the control system designed in this paper meet the needs,and the system has good engineering practicability.