RC 19494 (84719)
Modeling projects are often faced with a large parameter space that has to be explored in order to produce a set of performance measures representing the behavior of the systems under study. In this paper, we describe a software component that provides the analyst with the functionality to specify a design of experiments and execute a search algorithm over the resulting parameter space. The component invokes the associated simulation runs and compares the results to a goal to determine the solution. This component has been implemented as the run control mechanism in the RESearch Queueing Modeling Environment (RESQME). We demonstrate the use of this experimental run control and evaluation component for simple enumeration of the parameter space, interactive evaluation, as well as generalized rule-based control.
Gordon, Robert F. Ph.D.; Chang, Kow C.; and MacNair, Edward A., "Rule-Based Run Control and Evaluation for Simulation" (1994). Faculty Works: Mathematics & Computer Studies. 14.