An experimental design strategy for designing robust systems using discrete-event simulation

Abstract

Well designed experiments can dramatically improve the statistical accuracy of simula tion output data and facilitate the statistical analysis. In this paper we propose an experimental design strategy for designing robust systems using discrete-event simulation. The proposed strategy is motivated by Genichi Taguchi's strategy for improving product and process quality in manufacturing. Robust systems are systems that not only satisfy performance criteria, but also are not sensitive to uncontrollable environmental conditions or noise in the system's environment. We demonstrate the proposed experimental design strategy with a jobshop example and discuss the potential benefits of employing such a strategy for designing robust systems in general.

Monte Carlo Techniques for Parametric Finite Multidimensional Integral Equations

“…Most engineering system designs [13,26,44], such as product, process, and service design, involve a framework for arriving at a target value for a set of experiments, which may include Monte Carlo experiments. A random quality loss function L(Z i ) for a given system can be expanded in the neighborhood of the target value τ as follows:…”

A response to Kleij nen's note on Wild and Pignatiello's experimental design strategy

“…We would also like to thank the editors of Simulation for providing this forum for an exchange of ideas.The primary intention of our paper [5] was, (1) to focus on the importance of designed experiments in simulation work and to encourage simulationists to focus on this important activity, and (2) to bring to the attention of simulation researchers and practitioners the importance of designing for robustness.We did not intend to imply that the specific technique we presented is the only technique for finding robust system designs. We believe, however, that the technique we proposed provides an intuitively appealing argument for designing for robustness.…”

“…The primary intention of our paper [5] was, (1) to focus on the importance of designed experiments in simulation work and to encourage simulationists to focus on this important activity, and (2) to bring to the attention of simulation researchers and practitioners the importance of designing for robustness.…”

“…It can be shown [1] that the combined array layout can be used to reduce the total number of runs. The authors have been investigating this concept to devise a more economical experimental design strategy for robust design in simulation work than the one we introduced in [5].…”

“…We chose to use two-level environmental parameters for the jobshop example in our paper [5] for purposes of clarity of illustration of our proposed experimental design technique.…”