作者: Zhenyu Kong , Ramesh Kumar , Suren Gogineni , Yingqing Zhou , Jijun Lin
关键词:
摘要: Dimensional control has a significant impact on the overall product quality and performance in large complex multi-station assembly systems. From measurement data, way to identify root causes for variation of Key Product Characteristics (KPCs) is one most critical research topics dimensional control. This paper proposes new approach multiple fault diagnosis process by integrating multivariate statistical analysis with engineering model. Based product/process information, using state space model, set patterns are developed, which explicitly represent relationship between error sources KPCs. The vectors these form an affine system. Afterwards, Principal Component Analysis (PCA) applied conduct orthogonal diagonalization data. Thus, data can be easily projected axes Whereby, significance each pattern shall estimated accurately. Finally, few case studies also provided validate proposed methodology.Copyright © 2005 ASME