| Simulation optimization provides a structured approach to system design and configuration in the case where analytical expressions for quantities of interest are unavailable. This research focuses on the development of a new simulation optimization technique applicable to systems having multiple performance measures. The aim of this research is to incorporate a simulation end user's preference towards risk and uncertainty into the search process for the best decision alternative, a feature lacking in current techniques. Automation of the optimization procedure is a necessity. Therefore, this thesis proposes a simulation optimization method that involves a preference model, which is derived under axioms consistent with Von Neumann and Morgenstern utility theory along with three additional axioms.; The proposed preference model allows for a more manageable assessment procedure than what is currently available for the multiple attribute utility (MAU) model for problems that are non-monetary in nature and is shown in empirical studies to provide a better goodness of fit to a simulation end user's preference structure than a MAU model. The proposed simulation optimization method is evaluated against two simulation optimization methods with embedded deterministic, multiple criteria decision making strategies. It is shown to obtain significantly better solutions in multiple types of experimental settings that involve the distributions of the simulation performance measures to be normal. |
