Two foraging algorithms for a limited number of swarm robots

Abstract

Foraging behavior of ants can be beneficial when used in robotic applications that involve traveling between two points, such as harvesting, mining, and rescue robots. Several algorithms have been developed by imitating this behavior, but most of them require a large number of robots to perform efficiently. This paper proposes two novel algorithms that can imitate swarm behaviors using a limited number of robots. Both methods are constructed as decentralized systems that can function in unfamiliar environments. Virtual pheromone field (VPF) uses a sampling-graph based method to construct virtual pheromone trails that attract other robots. Sampling-graph based foraging (SGF) employs connected graphs to mitigate the effect of random movement. These two algorithms are simulated and compared with uncooperative (UC) swarm robots. Both proposed methods increase the efficiency and robustness of the swarm, while SGF provides the best results in the benchmark.