Simuleringsmetoder innen operasjonsanalyse : en oversiktsstudie

Abstract

Simulation is an important method in the field of operational research (OR). This report gives an introduction to simulation, explaining the terminology in use. Four main categories of methods are described: a) Static Monte Carlo simulation; b) System dynamics; c) Discrete event simulation; d) Agent-based simulation. For each of these methods, the main principles are described, examples of use are given and tools are listed. The report also discusses when it is appropriate to chose simulation as the solution method when faced with a problem. Analytical methods are generally preferable, but not always available, and simulation is then often the best option. Having chosen simulation as the analysis approach, the report gives guidelines on how to choose the appropriate simulation method. This decision largely depends on three main factors: a) the role of time; b) the presence of stochastic events; c) the level of system abstraction. The report describes and employs a simple military scenario for the purpose of illustrating examples of problems that can be addressed with the different simulation methods. This report is part of a series of surveys of OA methods conducted by project GOAL at FFI.