Front End Analytics and Caterpillar to Jointly Present Novel Statistical Model Validation Method at the ASME Verification & Validation (V&V) Symposium to be held in Las Vegas, NV on May 18-20, 2016
Front End Analytics and Caterpillar will present the paper titled “SIMULATION TIME HISTORY VALIDATION VIA MULTIVARIATE HYPOTHESIS TESTING OF FOURIER COEFFICIENTS” at the 2016 ASME Verification and Validation Symposium to be held in Las Vegas, NV from May 18-20, 2016.
Here is an excerpt from the Abstract:
“Some of the most common outputs of product development simulations are time histories that are relevant to some aspect of product operation. Quite often, these time histories in a physical test or customer usage environment exhibit considerable variability while at the same time maintaining some qualitative similarities. From a simulation validation perspective, the natural question to ask is whether it is plausible that a candidate set of simulation time history(ies) belongs to the same population from which a collection of reference (physical test) time histories were randomly sampled. The very nature of this question is suggestive of the concept of hypothesis testing. In this work, all curves in the candidate and reference sets were expressed via discrete Fourier series, and then the chosen coefficients were compared via a multivariate hypothesis testing process. Two different classes of problems are considered.”
To learn more about Front End Analytics, go to: www.feasol.com
To learn more about Caterpillar, go to: www.caterpillar.com
To learn more about the ASME V&V Symposium, go to: https://www.asme.org/events/vandv/