Design for Six Sigma through collaborative multiobjective optimization

Abstract

This paper presents a global methodology for designing product for Six Sigma. 
First, we combine a feasibility-modeling technique with an interactive multiobjective 
algorithm taking into account the decision maker's preferences (IMOP) to generate 
several Pareto-optimal solutions that maintain a probability of constraint satisfaction. 
These solutions are called reliable Pareto-optimal solutions. The solutions found by 
the algorithm fulfill as much as possible the decision makers' requirements. 
Second, we develop a procedure for choosing a solution for implementation from among 
the reliable Pareto-optimal solutions generated by the algorithm. This procedure is 
based on the robust design and philosophy of Six Sigma. Finally, the critical 
characteristics are identified to help the managers develop the manufacturing system 
and its related control plans in order to achieve quality products. The proposed 
methodology is applied to vehicle crash-worthiness design optimization for side 
impact with structural weight and front door velocity under side impact as objectives.