Gift value algorithm in multi-robot systems to promote cooperative coalitions

摘要

Effective cooperation among a group of robots is a key problem in multi-robot systems. The process of achieving cooperation can be modelled as coalition formation among a group of multi-robots. This paper presents a gift value algorithm based on environment knowledge, history experience and credit. The algorithm encourages selfish robots to form cooperative coalitions to optimize completion of tasks to get rewards using electoral and negotiatory methods based on gift value. The gift value algorithm is applied in robosoccer to improve test results.