基于自学习模糊Petri网的知识化制造系统采购预测  

Purchase forecasting in knowledgeable manufacturing systems based on self-learning fuzzy Petri nets

作　　者：孟宪刚[1] 严洪森[1] 

机构地区：[1]东南大学复杂工程系统测量与控制教育部重点实验室,南京210096

出　　处：《控制与决策》2009年第3期371-376,共6页Control and Decision

基　　金：国家自然科学基金项目(60574062);国家863计划项目(2007AA04Z112)

摘　　要：对模糊Petri网进行改进,使其增加自学习能力,即自学习模糊Petri网(SFPN).提出了自学习模糊Petri网模型知识库的建立方法,通过构造SFPN模型知识库,建立并保存现有产品的SFPN模型,开发新产品或进行新的决策时调出并进行修正后作为新产品模型,通过较短时间和少量样本的自学习训练,便可用于新产品的预测或决策.最后通过采购预测实例验证了该方法的有效性.Fuzzy Petri nets are improved by adding the ability of self-learning, namely self-learning fuzzy Petri nets （SFPN）. The method of building the self-learning fuzzy Petri nets model knowledge base is proposed. SFPN models of existing products are established and kept by building the SFPN model knowledge base. While a new product is developed or a new decision is made, the SFPN model of existing product can be read, modified and used as that of new product, which can be used to forecast or make a decision on new product through self-learning and self-training with less time and a few of samples. Finally, an application of forecasting real purchase shows the effectiveness of the method.

关 键 词：知识化制造 模糊PETRI网 知识库 

分 类 号：F227[经济管理—国民经济]