POLYNOMIAL METAMODELING AND TAGUCHI DESIGNS IN SIMULATION WITH APPLICATION TO THE MAINTENANCE FLOAT SYSTEM

This paper presents a new approach to modeling maintenance float systems via simulation. A regression metamodel is developed and applied in a decision framework. In order to derive the metamodel, Taguchi's design technique is used to identify the appropriate design for the controllable factors. Each factor is studied at three levels in order to consider both its linear and quadratic effects on the dependent variable, the average equipment utilization (EU). A polynomial regression method is used to determine the significant factors. The procedure offers an expedient method for solving complex production problems.