Research On Multi-Agent Negotiation Of Personalized Product Supply Chain Based On Q-Learning

With the development of domestic and foreign economy,people's consumption level has been continuously improved,and the concept of consumption has undergone a "quality" change.Consumers' demand for products has not only been limited to the products themselves,but also whether the product has "personalization" has become the focus of consumers' attention.In addition,with the rapid development of information technology such as the Internet,big data,cloud computing,artificial intelligence,etc.,the market environment faced by enterprises has undergone tremendous changes.The competition between enterprises is gradually transformed from product-and resource-oriented competition to customer-and informationoriented competition,and personalized supply chains are created..However,because the enterprises in each node of the personalized supply chain have their own interests,it is inevitable that there will be conflicts.How to ease conflicts and improve the efficiency of negotiation has become the key issue in the research of personalized supply chain.Based on the above background,this paper firstly summarizes the related theories of Agent and Multi-agent technology,personalized product supply chain and negotiation.Secondly,the main status quo of Multi-agent negotiation research is analyzed.On this basis,the efficiency of Multi-agent negotiation is discussed.Furthermore,for the negotiation of personalized product supply chain,this paper applies the Q-learning algorithm with autonomous learning ability to the research of negotiation problem,so as to construct the single-objective and multi-objective negotiation model of personalized supply chain Multi-Agent based on Q-learning algorithm.Finally,using MATLAB to simulate the two negotiation models,verify the feasibility of the model by an example.Considering in the reality,product integrators and customers are not simply accepting or rejecting a quote,but there is a degree of acceptance,That is,the unequal quotation of the negotiating participants will also promote the negotiation success,so this paper adopts a fuzzy algorithm to determine the final transaction price of the negotiation.The research characteristics of this paper are mainly as follows:1.Analyze the source and development status of personalized product supply chain,and propose the operation process of personalized product supply chain;2.Combining Q-Learning method with system,applying it to the negotiation of personalized supply chain,establishing a Multi-agent negotiation model of personalized supply chain based on QLearning,exploring and analyzing the optimal strategy and influencing factors to improve the negotiation efficiency of individualized supply chain and reduce the negotiation cost;3.Applying the Fuzzy decision method to the deadline for final offer of negotiation,so that the negotiation model is closer to reality.The research in this paper not only provides effective models and methods for the research of Multi-agent negotiation mechanism of personalized product supply chain,but also provides theoretical support for the practice of related enterprises.