| Robust design is an important method for reducing variation and improving product quality by selecting the levels of the controllable factors to minimize the effects of controllable noise variables. This is done by using experimental design and response surface methodology. Utilizing these methods, we develop new strategies for robust design for products and processes that have multiple responses of interest.;A strategy that can be used to solve general multiple response robust design problems would be a valuable contribution to the literature. Such a strategy can be developed using the concept of desirability functions to determine the settings of the design factors, not only to get the average performances on target but also to minimize variability around the target values.;To show that the strategies that we have developed can be successfully used, two examples (a stochastic inventory system and a traffic light system) are constructed to demonstrate and investigate there strategies in application areas. The performances of the new strategies are also compared and evaluated using these examples.;As a result of the inventory model example, the one-step direct strategies were chosen as being the best strategies. The best strategies for the traffic light system were generally similar to those of the inventory model. |
