Design And Implementation Of Welding Robot System For Thick Plate Parts

With the increasing demand of large equipment in related industries,the traditional manual welding method has gradually failed to meet the production demand.In this paper,cutter head parts of shield machine are taken as welding objects.Through the study of machine vision,digital image processing,welding path planning,automation integration technology and other aspects,this paper designed and implemented a welding robot system for thick plate parts based on laser vision.The system has completed the automatic identification of deep welding seam and the independent planning of welding path,which has improved the automatic level of welding of thick plate and increased production efficiency.First,combined with the welding characteristics of thick plate parts,the function of the system is analyzed,and the system is divided into four parts:welding planning module,integrated control module,motion execution module and welding operation module.Then,the paper determines the overall layout and architecture of the system,at the same time,the paper establishes the workflow of the system and finally completes the design of the overall scheme of the system.Second,according to the module division of the system,the paper analyzes the functions and characteristics of the hardware equipment required by the system,and completes the design or selection of the system hardware equipment.At the same time,combined with the system layout architecture,the paper builds a reasonable hardware communication network and completes the design of the system hardware platform.Third,the paper focuses on the analysis and design of the main software functions of the system.This paper compares the processing effects of different algorithms to complete the image filtering,segmentation and extraction of the center line of the weld.By improving the slope analysis method,angle analysis method is used to complete the accurate extraction of the feature points of the weld.Then,the spatial coordinate transformation of the robot and the camera calibration principle are studied,and a calibration method is derived to complete the hand-eye calibration of the system without the help of camera internal parameters.And then,based on the multi-layer and multi-channel path planning model and the idea of off-line programming,a layered welding method is proposed to reduce the influence of accumulated errors in the welding process and complete the adaptive welding of the thick plate.At the same time,the paper analyzes the requirements of data communication between each module equipment,selects and formulates reasonable software communication protocol,and improves the communication network of the system.Finally,according to the main functions of the system,a simple and intuitive human-computer interface is designed and developed to realize the state monitoring of the system.Fourth,according to the overall scheme and the design of software and hardware platform,the system is built and tested.The paper firstly tests the key parts of the system,such as vision module,welding planning and welding effect,and then tests the system function as a whole to verify the feasibility of the system.Through the test,the overall welding qualification rate of the system reaches 96.7%,and the welding forming is good,which meets the basic requirements of the welding work of medium thick plate.The system has been initially put into the welding production of part of shield machine tool head parts and is expected to be further popularized in the field of thick plate welding.