New simulation output analysis techniques: a statistical process control approach for estimating the warm-up period

ABSTRACT

One means for dealing with initialization bias in simulation experiments is to implement a warm-up period. This requires the correct estimation of the initial transient. A new method for determining the warm-up period, based upon the principles of statistical process control (SPC), is described. The method is tested on empirical data from a simulation model that has been used in a real-life study. In comparing the results to those from two commonly used warm-up methods, it appears that the SPC method performs well. The strengths and weaknesses of the approach are discussed.

REFERENCES

• 1: Banks, J., J. S. Carson, and B. L. Nelson. 1996. Discrete Event System Simulation. 2nd ed. Upper Saddle River, NJ: Prentice-Hall.
• 2: Bissell, D. 1994. Statistical Methods for SPC and TQM. London: Chapman & Hall.
• 3: Fishman, G. S. 1971. Estimating sample size in computer simulation experiments. Management Science 18:21--37.
• 4: Fishman, G. S. 1973. Concepts and Methods in Discrete Event Digital Simulation. New York: Wiley.
• 5: Fishman, G. S. 1978. Grouping observations in digital simulation. Management Science 24:510--521.
• 6: Gafarain, A. V., C. J. Ancker, and T. Morisaku. 1978. Evaluation of commonly used rules for detecting steady-state in computer simulation. Naval Research Logistics Quarterly 25:511--529.
• 7: Goldsman, D., L. W. Schruben, and J. J. Swain. 1994. Tests for transient means in simulation time series. Naval Research Logistics Quarterly 41:171--187.
• 8: Geoffrey Gordon, System Simulation, Prentice Hall PTR, Upper Saddle River, NJ, 1977
• 9: Jackway, P. T., and B. S. DeSilva. 1992. A methodology for initialization bias reduction in computer simulation output. Asia-Pacific Journal of Operational Research 9: 87--100.
• 10: Kelton W. D., and A. M. Law. 1983. A new approach for dealing with the startup problem in discrete event simulation. Naval Research Logistics Quarterly 30:641--658.
• 11: Averill M. Law , David M. Kelton, Simulation Modeling and Analysis, McGraw-Hill Higher Education, 1999
• 12: MacCarthy, B. L., and T. Wasusri. 2001. Statistical process control for monitoring scheduling performance - addressing the problem of correlated data. Journal of the Operational Research Society 52:810--820.
• 13: Nelson, B. L. 1992. Statistical analysis of simulation results. In The handbook of industrial engineering, 2nd ed, ed. G. Salvendy. New York: Wiley.
• 14: Krzysztof Pawlikowski, Steady-state simulation of queueing processes: survey of problems and solutions, ACM Computing Surveys (CSUR), v.22 n.2, p.123-170, June 1990 [doi>10.1145/78919.78921]
• 15: Robinson, S. 1994. Successful simulation: a practical approach to simulation projects. Maidenhead, UK: McGraw-Hill.
• 16: Roth, E. 1994. The relaxation time heuristic for the initial transient problem in M/M/k queueing systems. European Journal of Operational Research 72:376--386.
• 17: Runger, R. C., and T. R. Willemain. 1996. Batch-means control charts for autocorrelated data. IEEE Transactions 28:483--487.
• 18: Schruben, L. W. 1982. Detecting initialization bias in simulation output. Operations Research 30:569--590.
• 19: Schruben, L. W., H. Singh, and L. Tierney. 1983. Optimal tests for initialization bias in simulation output. Operations Research 31:1167--1178.
• 20: Vassilacopoulos, G. 1989. Testing for initialization bias in simulation output. Simulation 52 (4):151--153.
• 21: Stephen S. Lavenberg, Computer Performance Modeling Handbook, Academic Press, Inc., Orlando, FL, 1983
• 22: K. Preston White, Jr., An effective truncation heuristic for bias reduction in simulation output, Simulation, v.69 n.6, p.323-334, Dec. 1997 [doi>10.1177/003754979706900601]
• 23: Wilson, J. R., and A. B. Pritsker. 1978. A survey of research on the simulation startup problem. Simulation 31:55--59.
• 24: Yucesan, E. 1993. Randomisation tests for initialization bias in simulation output. Naval Research Logistics Quarterly 40:643--663.