Sensitivity analysis of a manufacturing workstation using perturbation analysis techniques

Sensitivity analysis of performance measures, like through-put or utilization, with respect to system parameters, such as service time parameters or buffer sizes, leads to improving and optimizing the performance of a given system. For a sample system from the world of manufacturing, we discuss various methods for estimating the sensitivity of the steady-state through-put. We compare the application of the finite difference estimator (where the sensitivity is estimated with the help of two independent simulation experiments) with the application of the infinitesimal and finite perturbation analysis (where the sensitivity is estimated with only one simulation experiment). Numerical examples are given to illustrate the practical aspects of the application of these methods.

[1]  Raif O. Onvural,et al.  Survey of closed queueing networks with blocking , 1990, CSUR.

[2]  Gianfranco Balbo,et al.  Computational algorithms for closed queueing networks , 1980 .

[3]  J. Kleijnen Statistical tools for simulation practitioners , 1986 .

[4]  B. Heidergott Infinitestimal Perturbation Analysis for discrete event systems with discrete lifetime distributions , 1994 .

[5]  Xi-Ren Cao,et al.  Convergence properties of infinitesimal perturbation analysis , 1988 .

[6]  Debasis Mitra,et al.  Control and coordination policies for systems with buffers , 1989, SIGMETRICS '89.

[7]  Paul Glasserman,et al.  Structural Conditions for Perturbation Analysis Derivative Estimation: Finite-Time Performance Indices , 1991, Oper. Res..

[8]  Xi-Ren Cao,et al.  Perturbation analysis of discrete event dynamic systems , 1991 .

[9]  R. Suri,et al.  Perturbation analysis: the state of the art and research issues explained via the GI/G/1 queue , 1989, Proc. IEEE.

[10]  M. Johnson,et al.  Sensitivity analysis of serial transfer lines using finite perturbation analysis , 1989 .

[11]  R. Rubinstein,et al.  Convergence of perturbation analysis estimates for discontinuous sample functions: a general approach , 1988, Advances in Applied Probability.

[12]  Paul Glasserman,et al.  Gradient Estimation Via Perturbation Analysis , 1990 .

[13]  W. J. Gordon,et al.  Closed Queuing Systems with Exponential Servers , 1967, Oper. Res..

[14]  P. Glasserman,et al.  Strongly Consistent Steady-State Derivative Estimates , 1991, Probability in the Engineering and Informational Sciences.

[15]  Christos G. Cassandras,et al.  Infinitesimal and finite perturbation analysis for queueing networks , 1982, 1982 21st IEEE Conference on Decision and Control.

[16]  Xi-Ren Cao,et al.  First-Order Perturbation Analysis of a Simple Multi-Class Finite Source Queue , 1987, Perform. Evaluation.