Font Size: a A A

White Body Fabrication Error Data Processing Method Comparison Research

Posted on:2016-07-18Degree:MasterType:Thesis
Country:ChinaCandidate:Z X BuFull Text:PDF
GTID:2272330464467860Subject:Vehicle Engineering
Abstract/Summary:PDF Full Text Request
Among all the factors affecting the quality of automobile manufacture, white body welding quality plays a crucial role. Therefore, it’s important to monitor and improve the quality of white body welding. At present, the diagnosis of fabrication error of the white body has been limited to a kind of method used independently, and does not have a comprehensive regional identification method.The source of this article is based on "2 mm project", in this paper introduced the causes of the fabrication error of the white body, and the development situation of "2 mm project". In order to find a comprehensive and effective error identification method, this article simply introduces the basic theorety of wavelet analysis, clustering analysis and principal component analysis. On this basis, use the white body fabrication error data do wavelet filtering and the principal component filtering respectively; And then in a software of the Autocad-second- development, finish the clustering analysis, and show the clustering result in Autocad.In this article, to the primary data of the white body fabrication error of three coordinate data do clustering analysis by the principle of clustering analysis, according to the results of the clustering recognition fabrication error area. Then use the trend of the wavelet filter do clustering analysis, the number of measuring points increased in the clustering results, the structure of the area to be more perfect, and consistent with the practical fabrication process of the white body, identify a new error area. Secondly, to the principal component analysis filter do clustering analysis,the number of measuring points increased in the clustering results, the structure of the area to be more perfect, and consistent with the practical fabrication process of white body, identify a new error area.Finally, to the trend of the wavelet filter and principal component filter finished the clustering analysis respectively, comparate the clustering results. Found that to the two types of filter do clustering analysis, the clustering results are basically similar, the construct of area is very complete, consistent with the practical fabrication process of the white body are, and identifythe same error area, contribute to the effective control of fabrication error of the white body.
Keywords/Search Tags:white body, Error area, Clustering analysis, Wavelet analysis, Principal component analysis
PDF Full Text Request
Related items