| The importance of the supply chain distribution system is increasing with the improvement of market dominance.The popularity of electronic trading provides new ideas to solve the problem of low efficiency in the supply chain,and the search for automated negotiation mechanisms has gradually become a trend.The problem of multilateral negotiation in the supply chain is more common than the bilateral negotiation problems.The original bilateral single attribute negotiation model can not restore the real scene of multilateral negotiation,especially when there are many constraints between multiple parties.In order to solve the problem of multilateral negotiation,this work proposes a multiattribute automatic negotiation model based on improved non-dominated sorting genetic algorithm NSGA-Ⅲ.At the same time,this work designs a multi-threaded negotiation process-In the multi-dimensional space,the Pareto font is solved by the algorithm,and then it is decomposed into multiple pairs of bilateral negotiations.In order to help the model to be more suitable for the complex and changeable negotiation environment,this work uses the multi-strategy selection theory to simulate the negotiation process.After the model is established,many comparative experiments are carried out to verify that the optimization of the algorithm and negotiation process can promote the negotiation agent to reach an agreement with the opponent faster in the negotiation process and reach a win-win situation.Finally,this work verifies the practicability of the model in the actual case by simulating the distribution problem of L company in the supply chain link.L company now needs to cooperate with the three major e-commerce platforms to sell products.Through multiple rounds of negotiations on the three attributes of operation cost,delivery quantity,and settlement cycle,the parties finally effectively give the solution,which not only ensures the interests of all parties but also protects the private information of all parties and improves the operational efficiency of the distribution system. |