A Monte Carlo Based Procedure for Analyzing Discrete-Time, Nonstationary Simulation Responses Using Classical Time Series Models

Author

Brandão, Rita Marques ; Nova, Acácio M O Porta

Author_Institution

Dept. de Mat., Univ. dos Acores, Ponta Delgada, Portugal

fYear

2009

fDate

20-25 Sept. 2009

Firstpage

6

Lastpage

11

Abstract

In this paper, we propose a procedure for analyzing discrete-time; nonstationary discrete-event simulation responses based on Monte Carlo integration and the use of classical ARIMA (or SARIMA) time-series models. The procedure is illustrated with an exploding single-server queue and a bounded cyclical traffic situation. Some conclusions and recommendations for further work are stated.

Keywords

Monte Carlo methods; autoregressive moving average processes; discrete event simulation; time series; Monte Carlo-based procedure; autoregressive integrated moving average; classical ARIMA; classical time series model; discrete-time analysis; nonstationary discrete-event simulation response; Analytical models; Discrete event simulation; Monte Carlo methods; Packaging; Predictive models; Sampling methods; Stochastic processes; Time series analysis; Traffic control; US Department of Transportation; Monte Carlo integration; discrete-event simulation; simulation metamodels; time-series models;

fLanguage

English

Publisher

ieee

Conference_Titel

Advances in System Simulation, 2009. SIMUL '09. First International Conference on

Conference_Location

Porto

Print_ISBN

978-1-4244-4863-0

Electronic_ISBN

978-0-7695-3773-3

Type

conf

DOI

10.1109/SIMUL.2009.15

Filename

5283981