To avoid unfair negotiation behavior in supply chain that consist of a manufacturer and many supplier, we introduce the theory of agent commitment, which take the acceptable proposal as a committed relationship. Base on this, we put forward agent commitment management mechanism to assist manufacturer pursuit the maximization utility. Under the background of supply chain collaboration procurement, we construct the multi-Agent concurrent negotiation model and deep study of the commitment management mechanisms and adaptive negotiation strategy.First, in view of the supply chain collaborative procurement, we build a complex multiple one-to-many concurrent negotiation framework. Besides, we design manufacturer agent promise management mechanism, which focus on reneging probability of the rival.Secondly, with regard to the poor adaptivity and inefficiency of one-to-many negotiation, we design adaptive negotiation strategy based on fuzzy reasoning. Timedependent negotiation strategy include three types: conservative, linear and conciliatory strategy. Each resource market conditions(such as supply and demand situation, the competitive level.etc) determine the corresponding agent negotiation status, so we also devide the negotiation status into there parts:advantage, balance and disadvantage.For each consultative status there is a negotiation strategy that can make the negotiation result best and suppliers choose one of them to negotiate. However, the manufacturer is not aware of resource market situation, we construct the parameter that can reflect supplier offer concessions, deriving the manufacturer in the market condition as manufacturer’s Agent of adaptive negotiation strategy using the theory of fuzzy reasoning.Thirdly, for more than one concurrent negotiation, we propose a utility orientation coordination strategy based on Radial Basis Function neural network within k-Nearest Neighbor(RBFNN-kNN),which is optimized by the k-Nearest Neighbor to improve Radial Basis Function neural network prediction accuracy.We use RBFNN-kNN to forecast the next round of consultation proposal, and select expected maximum utility offer according to the manufacturer’s Agent commitment management mechanism. At the same time we compare the expected maximum utility to the current proposal to decide whether to continue negotiation to ensure the order of multiple negotiation.Finally, we carry the simulation on the Matlab platform, and the simulation results show that the method is feasible in many areas of negotiation, so it’s possible to realize supply chain intelligent negotiation. |