83

Technical Research |Establishing Model Methodology

82

Technical Research |Establishing Model Methodology

One way to simplify the interactions between these entities may be to compare it with a simpler system. By looking back to Chapter 2.1, one can re-investigate the similarities between the crowd and water simulations. From this, it can then be observed that the crowd simulation, when dense enough, can arguably be described as a fluid. Fluid flows can be described in either the Lagrangian description—where each fluid particle is calculated as discrete particles—or the Eulerian description—where the fluid properties are only calculated at the boundary of set volumes.[5] (Fig. 2.2.2 - 3) While there are pros and cons to both methods, the Lagrangian methods will provide a more accurate model for non-continuum mass as well a more fluid simulation that would be more beneficial for crowd simulation. The downside to this is that extra computation is required to calculate each particle as more people are added into the system. However, by designing this model as a framework that can be scaled up, this limitation becomes less of an issue, especially with computation becoming faster with technological progression. However, humans are undeniably not as elementary as water particles. They do not simply flow, but have desires, motives, and goals. While it can be seen from the concert example that individuality starts to break down in larger crowds—and thus began to behave like the flow of water as the forces exerted on the individual people move in a direction that is determined by the crowd—it is still important to distinguish people in smaller groupings. Therefore, these rules at the bottom level will not only need to accommodate movement but also human behaviors. To create such a system, one must determine the logic behind one’s actions and look at ways to translate them into machine logic, essentially breaking them down to their fundamental elements.

The Lagrangian description calculates the position and velocity of the individual particles within the fluid

From “Descriptions of Fluid Flows,” accessed December 25, 2019, https://www.me.psu.edu/cimbala/Learning/Fluid/Introductory/descriptions_of_fluid_flows.htm.

The Eulerian description calculates the output velocities from the input velocities, in which the space inside the control volume is assumed to be completely filled as a continuous mass

From “Descriptions of Fluid Flows.”