SIMULATION IN MANUFACTURING - A REGRESSION METAMODEL APPROACH

The application of regression metamodels to simulation outputs is illustrated in this paper. With increasing interests in applying simulation to complex manufacturing problems, regression metamodels can greatly reduce the cost, time, and the amount of effort spent in conducting simulation. These models can also be generalized within the bounds defined for the problem's parameters. The use of a regression metamodel to conduct sensitivity analysis, application in "optimizing" manufacturing systems, and the validity of the model are illustrated. An example is given with a maintenance float problem. © 1990.