| Keywords [eng] |
simulation, multi-agent systems, ant, foraging, behavior, autonomous, agents, rules, collective intelligence, pheromone, decentralized coordination, scalable, biological systems, self-organization, swarm intelligence. simuliacija, daugiaagent˙es sistemos, skruzd˙el˙e, maisto paieška, elgsena, autonominiai, agentai, taisykl˙es, kolektyvinis intelektas, feromonas, decentralizuotas koordinavimas, mastelio keitimas, biologin˙es sistemos, saviorganizacija, spieˇciaus intelektas. |
| Abstract [eng] |
This study presents a quantitative evaluation of multi-agent systems through the lens of ant foraging behavior, implemented via an agent-based simulation framework using SimPy, a Python simulation tool. By modeling individual ants as autonomous agents interacting within a 2D grid-based ecosystem, we investigate how simple behavioral rules give rise to emergent collective intelligence. The framework allows systematic testing of environmental parameters, such as predator density, food distribution, and pheromone decay rates, to quantify their impact on foraging efficiency and colony resilience. The simulation provides insights into the resilience of decentralized coordination under hostile conditions and how predator pressure reshapes foraging behaviors. Finally, stresstesting under extreme conditions confirms the colony’s resilience, but also underscores performance degradation when multiple stressors coincide. Overall, the simulation confirms that simple local rules and effective communication can generate robust and scalable foraging strategies. This simulation not only demonstrates the principles of selforganization in biological systems but also serves as a versatile foundation for studying swarm intelligence in applied contexts such as robotics, AI or optimization algorithms. |
