基于多属性模糊Petri网的知识化制造系统产品需求预测

为强化模糊Petri网的知识表示能力, 提出了一种多属性模糊Petri网(AFPN), 即在模糊Petri网的基础上, 为每个库所及权值赋予与实际系统相对应的多个属性, 以简单模型结构完成具有多个属性的系统的建模．利用知识化制造系统的知识库, 对多属性模糊Petri网进行属性分解、属性合并以及训练更新, 并对不规则模型进行规则化, 可以在原有产品的AFPN模型基础上简单快速建立后续产品的AFPN新模型．在产品需求预测的实例中, 采用原材料和产品的各需求相关因素建立产品的需求预测AFPN, 实现了产品需求的快速预测．

Products demand forecasting in knowledgeable manufacturing systems based on attributes fuzzy Petri nets

For the sake of strengthening knowledge representation of fuzzy Petri nets,attributes fuzzy Petri nets(AFPN) are presented.On the basis of fuzzy Petri nets,places and weights are endowed with attributes which are related to actual system,so that the AFPN model of multiple attributes system is of simple model structure.AFPN models are stored in knowledge base of knowledgeable manufacturing systems,and irregular models are regularized after attributes decomposing,attributes merging,and model training.Thus,a new products processing AFPN model can be established simply and fast on the basis of original AFPN model.In the examples of products demand forecasting,demand forecasting AFPN model is established by using demand-related data of products and materials,and it can forecast products demands quickly.