A smart home resource management system for multiple inhabitants by agent conceding negotiation

摘要

This paper proposes a home resource management system for a smart home with multiple inhabitants. The system uses home ontology to covers the characteristics of available resources in smart homes. The ontology provides information of living space partition, energy consumption, and network bandwidth to detect resource usage conflict. Case-base reasoning is used to predict inhabitant resource requirement of smart homes. The system then uses BDI agent, blackboard systems, and conceding negotiation to find a common consensus plan of home resources for inhabitants. The BDI agent unfolds the inhabitants' information by belief, desire, and intension model. Blackboard system provides a platform for agent communication and resource usage conflict detection. Finally, conceding negotiation uses utility computation and conceder selection to find a common consensus resource allocation plan. The system is simulated in the sample smart home. It can provide an intelligent mechanism for efficient resource allocation.