Self-Organized Aggregation Based on Cockroach Behavior in Swarm Robotics

摘要

Aggregation has an important role for swarm robotic system, because it is at the basis of the emergence of various forms of many collective tasks. In this paper, we established the model inspired by the aggregation behavior in swarms of German cockroaches to realize the dynamic aggregation of self-organized and built difference equations to predict the aggregation of the robots. Aggregation emerges solely from local interactions between the individuals without leader and supervisor. The behavior of the individual depends on the probabilities to join or leave the aggregation which is in relation to local message. This model could make all the robots aggregate together and different aggregations can be formed in parallel. A method of improving the performance of aggregation was provided. Our work is based on a real robotic platform that is still under development. We present results and analysis of simulation-based experiments, which show that the model is robust and flexible.