Cascade Quality Prediction Method Using Multiple PCA+ID3 for Multi-Stage Manufacturing System☆

Abstract Quality prediction model, as the key to realize the real-time online quality monitoring process, has been developed using various data mining techniques. However, most of quality prediction models are developed in single-stage manufacturing system, where the relationship between manufacturing operation and quality variables is straightforward. Previous studies show that single-stage quality system cannot solve quality problem in multi-stage manufacturing system due to the complex variable relationships. This study is intended to propose a data mining method to develop quality prediction model which is able to deal with the complex variable relationships in multi-stage manufacturing system. This method, named Cascade Quality Prediction Method (CQPM), is developed by considering the complex variables relationships in multi-stage manufacturing system. CQPM employs the combination of multiple Principal Component Analysis and Iterative Dichotomiser 3 algorithm. A case study in semiconductor manufacturing shows that the prediction model that has been developed using CQPM is performed better in predicting both positive and negative classes compared to others.

[1]  S. Jack Hu,et al.  A Quality Prediction Framework for Multistage Machining Processes Driven by an Engineering Model and Variation Propagation Model , 2007 .

[2]  Fuli Wang,et al.  Quality Prediction of Strip Steel Based on Windows-Mean MPLS , 2010 .

[3]  Zhiqiang Ge,et al.  Utilizing transition information in online quality prediction of multiphase batch processes , 2012 .

[4]  I. Jolliffe.  Principal Component Analysis , 2002 .

[5]  Fuli Wang,et al.  Stage-based soft-transition multiple PCA modeling and on-line monitoring strategy for batch processes , 2007 .

[6]  Cunxi Xie,et al.  Fuzzy control for manufacturing quality based on variable precision rough set , 2004, Fifth World Congress on Intelligent Control and Automation (IEEE Cat. No.04EX788).

[7]  Xuejin Gao,et al.  Enhanced batch process monitoring and quality prediction using multi-phase dynamic PLS , 2011, Proceedings of the 30th Chinese Control Conference.

[8]  Xiao-Ping Guo,et al.  A stage-based quality prediction and control method for batch processes , 2005, 2005 International Conference on Machine Learning and Cybernetics.

[9]  Hyoungjoo Lee,et al.  Machine learning-based novelty detection for faulty wafer detection in semiconductor manufacturing , 2012, Expert Syst. Appl..

[10]  Richard E. DeVor,et al.  Statistical Quality Design and Control: Contemporary Concepts and Methods , 1992 .

[11]  Fuli Wang,et al.  Multi-PCA Models for Process Monitoring and Fault Diagnosis , 2003 .

[12]  Henry C. W. Lau,et al.  Development of an intelligent quality management system using fuzzy association rules , 2009, Expert Syst. Appl..

[13]  Costas J. Spanos,et al.  Fundamentals of Semiconductor Manufacturing and Process Control: May/Fundamentals of Semiconductor Manufacturing and Process Control , 2006 .

[14]  Fugee Tsung,et al.  Directional MEWMA Schemes for Multistage Process Monitoring and Diagnosis , 2008 .

[15]  H. Abdi,et al.  Principal component analysis , 2010 .

[16]  Shian-Shyong Tseng,et al.  A novel manufacturing defect detection method using association rule mining techniques , 2005, Expert Syst. Appl..

[17]  Costas J. Spanos,et al.  Fundamentals of Semiconductor Manufacturing and Process Control , 2006 .

[18]  Fan Jingjun,et al.  Alumina Production Operations Management Information System Based on Data Mining Technology , 2009, 2009 International Forum on Computer Science-Technology and Applications.

[19]  Francisco Herrera,et al.  Enhancing the effectiveness and interpretability of decision tree and rule induction classifiers with evolutionary training set selection over imbalanced problems , 2009, Appl. Soft Comput..