Agent-based model and system dynamics model for peace-keeping operations

Abstract

Military operations other than war (MOOTW) make up a large percentage of total military operations. Some common MOOTW operations are peacekeeping (PKO) and humanitarian assistance, and disaster relief (HADR). System dynamics (SD) uses a top-down approach that models high-level system behavior as compared to the use of agent-based modeling (ABM), which uses a bottom-up approach to generate system-level behavior through emergent behavior. In this work, SD and ABM were applied to model a food distribution scenario during the early phases of PK/HADR and the implementation process and results compared. The results were that large variations in food prices were observed as the time step and the integration technique were varied. Both SD and ABM, however, displayed similar emergent behavior in terms of crimes that occurred due to relative deprivation within the population. As an alternative to time step approximation, discrete event simulation (DES) may be used to implement the SD model through discretization of stocks or flows within the system and identifying events that change these quantities. The quantization of continuous variables in SD into discrete quantities may, however, introduce quantization errors. Emergent behavior seen in ABM can occur in SD through the interactions between equations. Due to the compactness of SD equations, it feels less intuitive to develop models using SD than it does to develop models using ABM.