A Mas-Based Negotiation Model And Its Application To Multiple Observation Collaborative Satellite Planning

A multiple agent system(MAS) architecture is constructed for multiple satellites collaborative planning system(MSCPS). A MAS-based negotiation model adopting extended contract net protocol(ECNP) is introduced to solve the dynamic task allocation for MSCPS in which the environment and capabilities of each Satellite Agent can dynamically change with time. In the negotiation model, Acquaintance's trusts based announcing bidding, adaptive bidding with swarm intelligence and multi-attribute decision based fuzzy evaluation bidding method are proposed respectively. Experiment results show the negotiation model can decrease communication cost and balance energy cost, hence solve task allocation effectively.