Optimizing Multiple Quality Characteristics

MC33.1 Run-to-Run Process Control Strategy Development Based on DOE

• Judy Jin; University of Michigan, Dept. of IOE, 1729 IOE Bldg., Ann Arbor, MI 48109-2117; jhjin@engin.umich.edu

Run-to-run control is a challenging problem, especially when control variables have interactions. We present a systematic hierarchical run-to-run control strategy based on DoE by using quality indices as control objectives, in-process signal features as response variables and control variables as factors. Examples for stamping process control are presented.

MC33.2 Trade-Off Issues in Multiresponse Experiments with the Consideration of Derringer-Suich's Desirability Function

• Young Jin Kim; Clemson University, Dept. of IE, Clemson, SC 29634; ykim@clemson.edu
• Byung-Rae Cho; Clemson University, Dept. of IE, Clemson, SC 29634; ykim@clemson.edu

Virtually all product quality is judged by multiple quality characteristics. Therefore, the use of a simultaneous optimization is inevitable to trade off these characteristics. We discuss the connections of fuzzy set to the Derringer-Suich desirability function and study how the desirability function can be extended to the framework of a multi-response experiment...

MC33.3 Some Approaches to Optimize Multiple Quality Characteristics

• Jai-Hyun Byun; Gyeongsang National University, Dept. of IE, 900 Gazwa-Dong, Chinju Gyeongnam, 660-701 , Korea; jbyun@nongae.gsnu.ac.kr
• Kwang-Jae Kim; Pohang University of Science & Technology, Dept. of Indust. Eng., San 31, Hyoja-Dong, Nam-Gu, Pohang, Kyungbuk, 790-784 , Republic of Korea; kjk@postech.ac.kr

We often have multiple quality characteristics to develop, improve and optimize industrial processes and products. It is not easy to optimize control factor settings where there are multiple quality characteristics. We suggest some approaches to optimize multiple characteristics and guidelines for the practitioners concern which approach to follow.

MC33.4 Evolutionary Operation for Multiple Quality Characteristics

• Jai-Hyun Byun; Gyeongsang National University, Dept. of IE, 900 Gazwa-Dong, Chinju Gyeongnam, 660-701 , Korea; jbyun@nongae.gsnu.ac.kr
• Yongyun Brian Kim; Gyeongsang National University, Dept. of IE, 900 Gazwa-Dong, Chinju, Gyeongnam, 660-701 , Korea; jbyun@nongae.gsnu.ac.kr

EVOP is a method for continuously monitoring and improving a full-scale production process to obtain an optimal operating condition while production is under way. We propose a procedure to simultaneously optimize multiple quality characteristics. Using a desirability function concept, we present 2 approaches to deal with multiple characteristic EVOP problems....