Black-box vs. White-box Models, Integrated vs. Cascade Simulations, and Multi-agent Simulation vs. Computational Fluid Dynamics

This article summarizes the concepts related to “simulation,” originally introduced by John von Neumann and considered one of the most highly visible scientific terms today. Simulations aim to model complex dynamic systems in the real world, allowing us to predict their spatiotemporal structure. Modeling and simulation are required to control any dynamic system, which in turn enables us to maximize the resource output from a system while mitigating constraints to ensure its sustainability. Simulation covers a wide range of dynamic systems from white-box to black-box models, depending on the transparency of each system. A physical system can be modeled using a white-box model because the governing equations are already given. However, the dynamics of a human decision process, or a human entity itself, is hardly treated as transparent, which inevitably leads to a black-box model. Another key idea related to simulation is “coupling,” which connects two or more different sub-dynamical systems using an integrated or cascade approach. One such important application is the merged technology of computational fluid dynamics and multi agent simulation. This paper presents concrete examples.