| With the intensification of enterprise competition,there is a trend of enterprises maintain both competition and cooperation in the transaction.Many enterprises have also realized that the purchase quantity of a single enterprise is small,which cannot form scale effects and has low bargaining power.Collaborative Procurement can enable enterprises to obtain more favorable prices,so as to enhance the competitiveness of enterprises and give full play to the advantages of industrial agglomeration areas.In the traditional supply chain,the cooperative enterprises are all independent enterprises with their own benefits.Both parties often consider maximizing their own benefits.When one party’s benefits are large,the benefits of the other party will be affected.Enterprises often have conflicts of interest in the process of cooperation.Negotiation is the basic means to solve the conflicts.Reasonable and effective negotiation model can help enterprises in industrial agglomeration areas to quickly resolve conflicts,so as to reach a plan with higher joint effectiveness in the negotiation,improve the overall competitiveness of the supply chain and achieve the purpose of win-win.With the support of JADE technology platform,this paper introduces a branch of distributed artificial intelligence(Multi-Agent technology)into the research of automatic negotiation,and formally defines the negotiation process of one-to-many and many-issue concurrent negotiation.This paper proposes a model of collaborative procurement negotiation in the agglomeration area of aviation equipment industry.The model judges whether the negotiation is terminated through the utility function,which changes the negotiation mode in the traditional negotiation that takes subjective judgment as the termination condition,It can standardize the negotiation process and improve the overall effectiveness of the negotiation.On the basis of this model,we have conducted in-depth research on negotiation strategies,study the negotiation strategy based on the opponent’s concession in detail,use particle swarm optimization to optimize the kernel function parameters σ and penalty factor C of support vector machine.We divide the historical negotiation data into training set and testing set,And use MATLAB for regression prediction,the prediction results show that PSO-SVM has higher prediction accuracy than traditional SVM algorithm,And We prove the feasibility of the adaptive negotiation strategy based on PSO-SVM with examples,and then prove the validity of the negotiation model used in this paper from the aspects of negotiation utility and negotiation efficiency.Finally,based on multi-agent negotiation model,adopting B/S architecture,Eclipse as programming tool,mysql as database,we designed a simple demonstration system of collaborative procurement. |
