Applying NURBs-Based Surrogate Models for Performance Forecasting in Manufacturing Systems

Manufacturing companies require systems that can respond to disruptions and be reconfigured quickly. The use of simulation to improve efficiency is particularly common; however, realistic and accurate simulations are computationally expensive. To save on computational expense, a facility manager can make use of a computationally efficient surrogate model that approximates the response of the simulation. This work implements a novel method of approximating throughput of a simulated manufacturing environment using Non-Uniform Rational B-splines (NURBs) as the basis for surrogate models. In three scenario studies, NURBs-based surrogates accurately approximate simulation outputs, with surrogate model query times ranging from 2 to 4 orders of magnitude faster than estimated evaluation times for corresponding simulations. These findings indicate that NURBs-based surrogates are a promising method of approximating manufacturing simulations for performance forecasting.Copyright © 2015 by ASME