中文摘要：

在保障生产运行的情况下,为达到合理占有库存资金的目的,提出了一种基于备件库存相关特性分类的备件库存管理方法。该方法将备件的库存方式进行分类,根据分类结果定义备件的分类树,将缺货成本、库存成本、采购成本、零件使用频率、备件供应情况和备件需求预测作为分类节点问题,使用模糊神经网络,确定其中需要多属性判断节点的值,最终依据备件分类树和库存策略表,实现对备件的分类库存管理。最后,给出了某特钢集团企业资源计划系统中该模块的软件实现。

英文摘要：

To keep the factory running smoothly with reasonable spare parts inventory fund, a new spare parts inventory management method was proposed based on the property classification of spare part inventory. Firstly, the spare part inventory styles were classified. Then, a decision tree was defined based on classification result with shortage cost, inventory cost, purchase cost, spare part utilization frequency, spare part supply condition arid spare part requirement forecast as the nodes of the tree. The value of the node was decided by fuzzy neural network if multi-attribute decision was needed. Finally, the classification management of spare parts inventory strategy ,was realized by using the decision tree and inventory strategy table. Software implementation of the model in a special steel enterprise＇s ERP system was also presented.