BIW assembly quality evaluation with variation of OCMM data and data-splitting error estimation

Hua Wang,Guanlong Chen,Ping Zhu
DOI: https://doi.org/10.1007/s00170-003-1790-z
IF: 3.563
2004-01-01
The International Journal of Advanced Manufacturing Technology
Abstract:Optical coordinate measuring machine (OCMM) has been widely used in body-in-white (BIW) assembly industry, but most of the studies associated with OCMM data are limited to detecting sudden changes of OCMM data from the control chart. The process of BIW assembly is a nonstationary time series. Leap, linear and periodical trends are three mean shifts of OCMM data. This paper introduces one simple and applicable method to evaluate the BIW assembly quality with OCMM data. Wavelet analysis is employed to separate the trend and variation in OCMM data. A practical case analysed by wavelet analysis proves the effectiveness of using variation of the OCMM data to evaluate the BIW assembly quality. The data-splitting error using wavelet analysis is estimated with Monte Carlo simulation .
What problem does this paper attempt to address?