A procedure for establishing a reference state in qualitative simulation of operational systems

Abstract

Identifies the need for a formal procedure to generate a reference state when conducting qualitative simulation studies. A procedure for generating a reference state is presented and justified. The procedure considers output performance indicators, steady state conditions, system capacity and model validity. Its application is illustrated using a simulation model for an order fulfillment process. The calibration process is necessarily iterative and subjective to some degree and in general does not generate a unique reference state. The impact of using different reference states in a simulation experiment is illustrated. In general the results are consistent when interpreted qualitatively in terms of direction, trend and order of magnitude. The approach is most applicable for qualitative simulation studies where comparative performance is being investigated and where precise numerical estimation is not required. However the procedure is also applicable to the study of real systems where data is not available.