摘要： 为提高供应链需求预测模型的精确度,利用联邦多Agent系统实现对Petri网供应链需求模型的预测。将企业需求供应链中的每个供应流程用一个Agent表示,以此模拟供应链的行为规则,根据Petri网寻找和复现供应链企业需求的实时变化规律,进而构建供应链的企业需求协同预测模型。与灰度值预测模型和支持向量机预测模型的仿真结果表明,该模型能提高企业供应链预测模型的计算效率。

关键词: 供应链, 需求预测, Petri网, 信息管理, 动态特性

Abstract: In order to improve the accuracy of the supply chain prediction model, this paper uses federal Multiple Agent System(MAS) to realize the prediction of Petri net supply chain model. It uses Petri net to search and recur real-time change rule of the supply chain enterprise needs, and builds the enterprise supply chain collaborative prediction model. Simulation results with the grey value prediction model and Support Vector Machine(SVM) prediction model show that the proposed model can improve the prediction efficiency of enterprise supply chain prediction model.

Key words: supply chain, demand prediction, Petri net, information management, dynamic characteristic

中图分类号: 

• TP391.9