Abstract
Data truncation is a commonly accepted method of dealing with initialization bias in discrete-event simulation. An algorithm for determining the appropriate initial-data truncation point for multivariate output is proposed. The technique entails averaging across independent replications and estimating a steady-state output model in a state-space framework. A Bayesian technique called Multiple Model Adaptive Estimation (MMAE) is applied to compute a time varying estimate of the output's steady-state mean vector. This MMAE implementation features the use, in paralle, of a bank of Kalman filters. Each filter is constructed under a different assumption concerning the output's steady-state mean vector. One of the filters assumes that the steady-state mean vector is accurately reflected by an estimate, called the “assumed steady-state mean vector”, taken from the last half of the simulation data. As the filters process the output through the effective transient, this particular filter becomes more likely (in a Bayesian sense) to be the best filter to represent the data and the MMAE mean estimator is influenced increasingly towards the assumed steady-state mean vector. The estimated truncation point is selected when a norm of the MMAE mean vector estimate is within a small tolerance of the assumed steady-state mean vector. A Monte Carlo analysis using data from simulations of open and closed queueing models is used to evaluate the technique. The evaluation criteria include the ability to construct accurate and reliable confidence regions for the mean response vector based on the truncated sequences.
Similar content being viewed by others
References
J.M. Charnes, Statistical analysis of multivariate discrete-event simulation output, PhD Dissertation, University of Minnesota, Minneapolis, MN (1989).
R.W. Conway, Some tactical problems on digital simulation, Manag. Sci. 10(1963)47–61.
G.S. Fishman, Estimating sample size in computing simulation experiments, Manag. Sci. 18(1971)21–38.
G.S. Fishman, Bias considerations in simulation experiments, Oper. Res. 20(1972)785–789.
G.S. Fishman,Principles of Discrete Event Simulation (Wiley, New York, 1978).
M.A. Gallagher, Identification of the initial transient in discrete-event simulation output using the Kalman filter, PhD Dissertation, Department of Operational Sciences, Air Force Institute of Technology, Wright-Patterson AFB, OH (1992).
M.A. Gallagher, K.W. Bauer and P.S. Maybeck, Initial data truncation for univariate output of discrete-event simulations using the Kalman filter, Technical paper, Department of Operational Sciences, Air Force Institute of Technology, Wright-Patterson AFB, OH (1992).
A.V. Gafarian, C.J. Ancker and T. Morisaku, Evaluation of commonly used rules for detecting “steady state” in computer simulation, Naval Res. Log. Quarterly 25(1978)511–529.
R.B. Howard, M,A, Gallagher, K.W. Bauer and P.S. Maybeck, Confidence intervals for univariate discrete-event simulation output using the Kalman filter, in:Proc. 1992 Winter Simulation Conf., ed. J. Swain and D. Goldsman, Arlington, VA (1992).
R.B. Howard, M.A. Gallagher, K.W. Bauer and P.S. Maybeck, Confidence interval estimation for univariate output of discrete-event simulations using the Kalman filter, Technical paper, Department of Operational Sciences, Air Force Institute of Technology, Wright-Patterson AFB, OH (1992).
W.C. Kelton and A.M. Law, A new approach for dealing with the startup problem in discrete event simulation, Naval Res. Log. Quarterly 30(1983)641–658.
A.M. Law and W.D. Kelton,Simulation Modeling and Analysis (McGraw-Hill, New York, 1991).
P.S. Maybeck,Stochastic Models, Estimation, and Control, Vol. 1 (Academic Press, New York, 1979).
P.S. Maybeck,Stochastic Models, Estimation, and Control, Vols. 2 and 3 (Academic Press, New York, 1982).
P.S. Maybeck and K.P. Hentz, Investigation of moving-bank multiple model adaptive algorithms, AIAA Guidance, Control, and Dynamics 10(1987)90–96.
P.S. Maybeck and R.D. Stevens, Reconfigurable flight control via multiple model adaptive control methods, IEEE Trans. Aerospace Elec. Syst. AES-27(1991)470–480.
R.K. Mehra, On-line identification of linear dynamic systems with applications to Kalman filtering, IEEE Trans. Auto. Contr. AC-16 (1971)10–21.
A.A.B. Pritsker,Introduction to Simulation and Slam II, 3rd Ed. (Wiley, New York, 1986).
S.M. Ross,Introduction to Probability Models (Academic Press, New York, 1985).
R.Y. Rubinstein and R. Marcus, Efficiency of multivariates in Monte Carlo simulations, Oper. Res. 33(1985)661–677.
T.J. Schriber and R.W. Andrews, ARMA-based confidence intervals for simulation output analysis, Amer. J. Math. Manag. Sci. 4(1984)345–373.
L.W. Schruben, Control of initialization bias in multivariate simulation response, Commun. ACM 24(1981)246–252.
L.W. Schruben, Detecting initialization bias in simulation output, Oper. Res. 30(1982)569–590.
H.J. Steudel and S.M. Wu, A time series approach to queueing systems with applications for modeling job-shop in-process inventories, Manag. Sci. 23(1977)745–755.
P.D. Welch, The statistical analysis of simulation results, in:The Computer Performance Modeling Handbook, ed. S.S. Laverberg (Academic Press, New York, 1983) pp. 268–328.
J.R. Wilson and A.A.B. Pritsker, A survey of research on the simulation startup problem, Simulation 31(1978)55–58.
J.R. Wilson and A.A.B. Pritsker, Evaluation of startup policies in simulation experiments, Simulation 31(1978)79–89.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Gallagher, M.A., Bauer, K.W. & Maybeck, P.S. Initial data truncation for multivariate output of discrete-event simulations using the Kalman filter. Ann Oper Res 53, 419–441 (1994). https://doi.org/10.1007/BF02136837
Issue Date:
DOI: https://doi.org/10.1007/BF02136837