A Unified Agent-Based Simulation Framework for Epidemic and Wildfire Propagation in Complex Adaptive Systems

Spreading phenomena in complex adaptive systems often exhibit similar dynamic characteristics despite originating from different application domains. This study investigates heterogeneous spreading processes using a common agent-based modeling methodology implemented in the AnyLogic simulation platform. Two representative models were developed: influenza transmission within a university population and wildfire propagation in a forest environment. Both models employ statechart-based agents, stochastic state transitions, and local interactions while preserving domain-specific characteristics. The influenza model incorporates mask usage, vaccination, and isolation as intervention strategies. Simulation results show that the investigated intervention scenarios reduced epidemic intensity compared with the baseline under the specific parameter settings considered, with peak infectious populations ranging from 44 to 70 individuals compared with 104 in the baseline scenario. The wildfire model investigates the influence of vegetation density, wind conditions, and burning duration, with burned trees ranging from 1 to 1,035 depending on environmental conditions. Comparative analysis demonstrated that both systems exhibit marked changes in propagation dynamics across the investigated scenarios, together with strong dependence on local interactions and stochastic processes. The successful application of a common agent-based modeling methodology across two distinct spreading phenomena highlights its consistency and transferability while supporting decision-making in public health and environmental risk management.