Utilizing Combinatorial Testing on Discrete Event Simulation Models for Sustainable Manufacturing

This paper describes how combinatorial testing using covering arrays can be implemented to optimize discrete event simulation models of manufacturing systems for measures of sustainability. Discrete event simulation models often have hundreds of parameters and many test values for each parameter. Generally the interactions between the parameter-values are not well understood; this can lead to sub-optimization of the system. Most optimization engines and software for discrete event simulation packages use full factorial designs, which require many runs and hence a lot of computation time. In this paper we introduce combinatorial testing using a test-suite generation tool called NIST-ACTS (National Institute of Standards and Technology - Advanced Combinatorial Test Suites) to dramatically decrease the number of runs required to detect the interactions and determine an optimal solution.