Research On Methodology And Technology Of Monitoring Body In White Welding Error

This paper derives from an actual project whose goal is to provide an automobile company to develop a body in white assembly and welding error monitoring software. The scope of study of this paper are the automotive industry "2mm Project" area, which includes monitoring the error and diagnosing the error source of the assembly and welding of automotive body-in-white(BIW).This paper has summarized the development of the issues faced by domestic and international situation, selected the plateform and determined the overall program, designed the main framework of the software, achieved the"program of displaying error through space and time"of assembly and welding error with the "error displaying module of time history " and "error displaying module of space"; provided the function of diagnosing error source with the development of the"Correlation Analysis module"and the"clustering module"; completed the union-debug; introduced serveral examples of engineering. The paper has also investigated the technology of developing with multi language and secondary development of AUTOCAD in-depth; used some cutting-edge technology within the field.The software, result of this paper, is developed basing on the secondary development platform AutoCAD and using C++, VB and EXCEL. Its database based on EXCEL and display the result within AutoCAD. The input data is the measuring data of CMM which is formated with dmo format. Firstly, it should be imported into EXCEL with the expected format and then push the result into the basic database of AutoCAD. Then , the software will create the error models with these data, and show then with in the AutoCAD.The innovation of this paper are mainly twofold: Firstly it has raised the"program of displaying error through space and time"and made it. As the error models are shown in 3D and could be rotated and zoomin or zoomout, we could observe the space status and history data in the same time. It has overcame a number of defects that the current general-purpose software that monitoring in 2D. The second is that it has designed clustering analysis algorithms that facing the process of welding, and provided an effective tool for diagnosing erorr.