Research On The Negotiation Strategy Of Multi-agent Supply Chain Production-marketing Synergy Based On The Fairness Perceptions

In many areas of the supply chain production and marketing, due to the manufacturers and retailers' different business objectives, cultural and operational methods, it is prone to generate conflicts. Only rapid and effective solutions for various contradictions can improve the cooperation relations between enterprises and ensure the effective operation of the supply chain. Meanwhile negotiations have become a key way; Agents can resolve and eliminate conflicts through negotiation and seek to maximize the overall interests of the supply chain. However, in real life, the enterprises on the basis of guarantee the overall interests, also pursue maximizing their own interests inevitably. Once they have a feeling of self-interest damaged; will have a negative attitude to cooperation, even quit to the entire supply chain seriously. So how to assess or reflect Agents' fairness perceptions influencing their behavior is particularly necessary and significance to the whole supply chain run orderly, stably, long-timely. Therefore, this essay is mainly about multi-Agents adaptive negotiation strategies under fairness perceptions.First, considering the impact of the Agents' fairness perceptions on their compromise behaviors, this essay gives the adaptive negotiating framework, model and process of the multi-Agents under fairness perceptions. The negotiation framework is establishing a multilateral negotiation on the basis of multi-issue, with the proposal submission, market transactions evaluation, negotiating strategy generation, three modules. The negotiation model is under the hypotheses of Agents' not completely self-interest, fairness perceptions and so on; builds with a six-tuple model. The negotiation process describes the Agent concession adjustment and optimal strategy generation procedure under the fairness perceptions.Secondly, from the two perspectives of fairness concerns and fuzzy fairness, using different methods and learning algorithms to represent fairness, the essay discusses the adaptive negotiation strategies under Agents' fairness perceptions. The first is regarding to the adaptive negotiation under equality fairness perception, and it considers the fairness concern behavior of Agent, uses the F-S unfair hating model by Ernst Fehr and Klaus M. Schmidt, optimizes Actor-Critic learning algorithm through RBF neural network to predict and adjust both sides Agent's concession magnitude; sequentially proposes a AC-RBF adaptive strategy. The second is more about Agent adaptive negotiation under fuzzy fairness, and it considers Agent perceptions of distributive fairness and procedural fairness, uses fuzzy logic to express the vague concept of fairness, combines fuzzy reasoning and BP neural networks, deduces Agent fairness acceptance degree through fuzzy network, reuse Actor-Critic learning algorithm to adjust the amplitude of Agent concession under the current negotiation status; sequentially proposes a AC-FNN adaptive strategy.Finally, the Matlab software is used to prove the two adaptive strategies' effectiveness. In the two-stage supply chain scenario with multi-issue negotiation of multilateral environment, this essay constructs the adaptive negotiation model under fairness perceptions, gives the initial numbers of negotiation issues and parameters, simulates the AC-RBF and the AC-FNN negotiation strategy separately, and verifies the results validity.