Comparison of Different Cue-Based Swarm Aggregation Strategies

Abstract

Abstract. In this paper, we compare different aggregation strategies for cuebased aggregation with a mobile robot swarm. We used a sound source as the cue in the environment and performed real robot and simulation based experiments. We compared the performance of two proposed aggregation algorithms we called as the vector averaging and naïve with the state-of-the-art cue-based aggregation strategy BEECLUST. We showed that the proposed strategies outperform BEECLUST method. We also illustrated the feasibility of the method in the presence of noise. The results showed that the vector averaging algorithm is more robust to noise when compared to the naïve method.

A review of swarm robotics tasks

“…In [25], the basic algorithm is enhanced with two modifications: robots change their walking velocity based on the local luminance, with higher velocity corresponding to darker areas, and increase their waiting time when stopped near a high number of neighbors. In [48,49], the aggregation area is signaled with a sound source, and robots are equipped with microphones to measure the sound intensity; in order to increase the aggregation efficiency, each robot has a set of microphones oriented at different directions, and when resuming walking moves toward the estimated direction of the sound source.…”

“…The temporal dimension in the aggregation task is taken into account in performance metrics that measure the speed at which a swarm achieves an aggregation target [20,25,48,49]; usually, such metrics are calculated as the time elapsed before a given percentage of robots forms an aggregate, starting from initial conditions where robots are positioned randomly in the arena.…”

Self-organization in aggregating robot swarms: A DW-KNN topological approach

“…Recently, one of the most successful cue-based aggregation models took inspiration from the collective behaviour of honeybees, which prefer gathering where the temperature is 36 • C. The BEECLUST model proposed by Kernbach et al [4] was the first algorithm that mimicked this behaviour; a gradual light source was used to generate clustering behavior in a swarm robotics system. It was proven to act robustly in many A c c e p t e d M a n u s c r i p t researches [19,20,21,22]. Further, different variations of the model have been suggested to increase the performance of the aggregation process.…”

“…For instance, Arvin et al [23] proposed a new aggregation algorithm in which a dynamic velocity and a comparative waiting time were introduced to the original BEECLUST model, which contributed to a significant improvement in the aggregation time. Furthermore, a comparison between the original BEECLUST algorithm and two modified versions -called the vector averaging algorithm and the naive algorithm -showed that both the vector averaging and naïve algorithms outperformed the original BEECLUST model, and revealed that noise has less impact in the vector averaging method than the naïve one [21].…”

Power-Law Distribution of Long-Term Experimental Data in Swarm Robotics

“…The results showed that both methods improve aggregation performance. In addition, the effects of turning angle has been studied in [6]. In this study, the performance of two proposed aggregation algorithms -vector averaging and naïve -was compared with BEECLUST.…”