JOURNAL OF SIMULATION, cilt.1, sa.1, ss.1-15, 2025 (SCI-Expanded)
Agent-based models (ABM) consist of agents, the environment, and the
interactions between them. In most agent-based simulation studies, the
focus is on the agents and their interactions. As one of the three main
components in the agent-based models, the environment holds great
potential to improve models’ performance and efficiency. However, in
general, the environment is usually treated implicitly, or in an ad hoc
manner. Increasing environmental support may improve the overall
modularity and reusability of the model. To this end, we present the
Permission-driven Interaction Model (PDIM) for agent-based models with a
centralized environment. PDIM is an interaction model that establishes
an environmental approval step for each action to ensure the enforcement
of environmental rules and transfers the execution of actions to the
environment. This approach provides a robust implementation of
environmental dynamics, increases the modularity and reusability of the
environment, and simplifies agent design since agents focus solely on
reasoning, rather than execution of their actions. In this paper, PDIM
is introduced and then demonstrated through a proof-of-concept defense
scenario, which is validated using conventional numerical methods to
ensure its reliability in constructing modular environments with robust
regulatory mechanisms while eliminating the need for ad hoc modeling.