Process Disturbance Identification using Independent Component Analysis with CART Approach

S.-Y. Huang, C.-C. Chiu (Taiwan), D.F. Cook (USA), and Y.-P. Luh (Taiwan)


: statistical process control engineering process control independent component analysis classification and regression tree


In the past few years many studies have been conducted about the integrated use of statistical process control (SPC) and engineering process control (EPC) because applying them separately cannot control the manufacturing process optimally. And most of these studies reported that the integration outperforms than using only SPC or EPC. Basically, the former aims to rapidly detect assignable causes and time points for abnormalities that Lake place during process', and the latter is a method in which input variables are adjusted against process outputs through a feedback control mechanism. However, though combining SPC with EPC call effectively detect time points when abnormalities occur during process, using these two conventional tools may give false alarms ill most cases if process data are auto-correlated to some extent. In this study to increase the accuracy of process disturbance identification we used independent component analysis (ICA) with classification and regression tree (CART) approach to identify and recognize shifts ill the correlated process parameters.

Important Links:

Go Back