Extended Contract Nets Application in Wood Drying Kiln

摘要

The article aims at the problems such as the traditional contract nets model's communication traffic is too heavy in the consultation process; task's partition is fixed; evaluation index is not comprehensive, proposes a contract nets improvement method based on the bidding threshold and evaluation function by combining the response threshold model of the ant colony algorithm. This method introduce fuzzy synthetic computation, finding the potential cooperation relations between node agents, bottom-up decomposing the complex task, effectively reduce the negotiation time and communication flows, improve the efficiency of the system. The thesis combines wood drying application feature, and puts forward the improved model's experimental simulation in wood drying kiln. The result shows that the improved method can effectively improve the performance of the Contract Net Protocol in large-scale multi-agent system. It provides technical basis for the realization of wood drying cluster control.