Mok-nskart: new procedure for estimating confidence interval of steady state simulation output mean

Abstract EN

Discrete-event simulation can be classified as either finite-horizon or steady-state.

Finite-horizon (terminating) simulation models ended at a specific time or by the occurrence of a specific condition.

On the other hand, steady-state (non-terminating) simulation operates (at least conceptually) into the indefinite future; and in this case the interest focus on long-run average performance.

Non-overlapping Batch Means (NBM) method is one of the most effective methods used for analyzing steady-state simulation output.

This paper presents a simulation procedure undertaken in the automotive industry.

Therefor it aims First, to develop a new procedure for steady-state simulation output analysis "Mok-Nskart", it is designed to deliver a confidence interval for the steady-state mean of a simulation output process.

Mok-Nskart can be considered as an extension of the method of NBM.

The present study aims secondly, to evaluate the performance of Mok-Nskart, by comparing it with other simulation analysis procedures—namely N-Skart and MSER-5.