Discrete event system simulation is something we do without necessarily calling it that. When I described it to Joe, he realized he was already doing it for the LRGV study. A discrete event system simulation is a simulation that evolves the state of a system over time; the “discrete event” part means that the system state remains the same between the instantaneous events that change it. This is in contrast to a continuous-time simulation such as might be represented by a system of PDEs, or a finite-element simulation that has no time-varying component at all. Discrete event system simulation is used in IE to study all manner of problems that don’t necessarily have analytical solutions, like factory planning, emergency department staffing, and supply chain management.
Dr. Deb Medeiros does a very good job teaching a graduate-level course on the subject, IE522. Since we tend to use our own handwritten simulations, the part of this course that’s probably of greatest interest to our group is the section on random numbers, which explores the statistical consequences of different approaches to using multiple streams of random numbers in a simulation experiment.