There and back again; Efficient foraging with heterogeneous Agents using Pheromones

Abstract

Foraging is a canonical task for swarm intelligence systems. Traditionally thistask has been preformed by a swarm of homogeneous agents, often with the helpof pheromones for indirect communication. Some researcher have experimentedwith how heterogeneous agents can achieve this task. Yet little research exist,combining morphologically heterogeneous agents with pheromone based foragingalgorithms. This is confirmed thought a structured literature review. C-SAF is anexisting foraging algorithm employing homogeneous agents, communication viapheromones. In this thesis, morphologically heterogeneous agents are introducedand combined with the core mechanics of C-SAF to creating a new algorithm,dubbed H-CAF. H-CAF is compared with C-SAF across different scenarios, usingwell established performance metrics from the literature.