Date of Award
Doctor of Philosophy (PhD)
Computational Modeling & Simulation Engineering
Modeling and Simulation
Modeling and simulation are applied in a great many methods across a variety of topics. Model developers and users alike have a professional duty to understand the complexities of the tools and methods they are used. Oftentimes, models that have been independently constructed and executed are used to inform one another for an analytic purpose and the compatibility of the models is not always addressed. In the literature, great attention has been paid to model validation. When using models constructively with one another, analysts must understand the bounds of model validity and ensure that the combination of models does not generate poor information. The literature reveals significant research on model interoperability and model composability. Special analytic cases of composability in multi-resolution modeling have also been examined in the available research. What is not available, however, is the ability to assess models’ ability to inform one another without violating the validation of either model. Therefore, the purpose of this research is to develop a risk of method to model composability. To develop this method, a macroscopic model, simulating large-scale transportation problems will be implemented. An available technique for Model Use Risk Methodology (MURM) will be applied to the macroscopic model to measure its appropriateness for use within its validated space. The model will be decomposed into atomic units of Objects and Processes. Next, a microscopic traffic model will be similarly decomposed into atomic units and be used to inform the macroscopic model. Applying model similarity techniques across the atoms of both models will yield an assessment of their compatibility of one another. The macroscopic model will be reassessed using the MURM. Changes in its risk-of-use score will be compared against the model elements’ similarity to derive a relationship between model similarity and its impact upon model use appropriateness.
Young, John B..
"Validation Risk Across Hierarchical Models"
(2020). Doctor of Philosophy (PhD), Dissertation, Computational Modeling & Simulation Engineering, Old Dominion University, DOI: 10.25777/d4vy-aw23