Collective Decision-Making Behaviors Control in Swarm Robotic Search

摘要

Taking decision-making on task allocation as an example of collective behaviors, the authors presented an algorithm based on the existing response threshold allocation model for swarm robotic search. To make swarm behaviors emerged, they built a dynamic adjusting mechanism and introduced it into the self-organized pattern formation framework. Here each robot selected its own desired target from its personalized task set autonomously first. The robots whose desired targets were the same formed sub-swarms to be ready for their respective common targets search in parallel. Then the level of robotic resource deployment within sub-swarm was measured independently as negative feedback by each member robot of sub-swarm. After then each member had to decide staying at or leaving from the current subswarm on the basis of evaluation. Accordingly, all sub-swarms adjusted their respective swarm size to get relative balanced robots distribution. Results from simulations reveal that collective decision-making behaviors can be controlled well.