Simulation models often have many input factors, and determining which ones have a significant impact on performance measures (responses) of interest can be a difficult task. The common approach of changing one factor at a time is statistically inefficient and, more importantly, is very often just incorrect, because for many models factors interact to impact on the responses. In this tutorial we present an introduction to design of experiments specifically for simulation modeling, whose major goal is to determine the important factors with the least amount of simulating. We discuss classical experimental designs such as full factorial, fractional factorial, and central composite followed by a presentation on Latin hypercube designs, which are designed for the complex, nonlinear responses typically associated with simulation models.
展开▼
