Research On Multi-Attribute Automatic Negotiation Model In One-to-Many Supply Chain Procurement

At present,there are many automatic negotiation models to solve the conflict problem in the distributed computing system,but only the single attribute negotiation of the price as the negotiation attribute does not satisfy the realistic situation.The multi-attribute negotiation is the image description of the real negotiation,and most of negotiation models does not well achieve a win-win result.In order to solve this problems,based on the multi-objective optimization of the evolutionary algorithm,this thesis proposes a negotiation model that can find a multi-attribute win-win solution,which can prevent the leakage of the negotiation agent’s own utility function.At the same time,this thesis designs a time-dependent strategy model that relies on time and behavior to help the two sides reach the final agreement more quickly in a series of win-win solutions.Finally,a large number of experiments confirm that the win-win negotiation model in this thesis has more reasonable results than the existing models in efficiency and utility promotion.The experiments also show that the model of this thesis is stable and effective when finding a fair win-win result,which is rarely solved in the existing models.Finally,this thesis verifies the practicality of the model by simulating the purchase problem of Company A.A company now needs to purchase a number of new bus engines,through the 10 suppliers to meet the requirements of choice,the two sides through a number of negotiations to reach a deal,from the 10 price,warranty period,the comprehensive effect of the delivery period,select the buyer The most effective supplier,as the ultimate supplier,to ensure the interests of enterprises,the results prove that bilateral multi-attribute multi-strategy selection theory automatic negotiation model in the enterprise utility and efficiency improvement is very effective.