Multi-response optimization approaches and their applications in parameter design for robust products
Course duration: 8 h
Designing robust products and processes that perform as desired consistently is a challenge for today’s manufacturers. Product and process parameter levels are tried to be selected in such a way that the product performs at its target level no matter what happens beyond the control of the designer. Since a product typically has several quality characteristics, finding the best levels of these parameters can be considered as a multi-response optimization problem. In this course, fundamental principles and approaches of multi-response optimization are discussed as they are implemented in robust parameter design of products and processes. First the single response robust parameter design problem and its solution approaches are introduced, then the case of multiple responses is covered. Development of empirical response surfaces is briefly discussed. Aggregation of multiple responses and performance measures used in the optimization are given special attention, including desirability functions. Different multi-objective optimization approaches to formulation and solution of robust parameter design problems are presented.
Prof. G?lser K?ksal
Place of employment: Department of Industrial Engineering, Department of Statistics, Institute of Applied Mathematics, Middle East Technical University, Ankara, Turkey
Spheres of researches: Data mining in quality improvement, product/process/service development through quality function deployment, design optimization (robust design), design for six sigma, six sigma, lean six sigma, integrated quality and production management, statistical process control