基于模糊聚类算法的定制厨柜零件分类制造技术  

Customized cabinet part classification manufacturing technology based on fuzzy clustering algorithm

作　　者：贾晔[1] 李军[1] 丁艳 

机构地区：[1]南京林业大学家具与工业设计学院,南京210037 [2]江苏省家具行业协会,南京210029

出　　处：《林业工程学报》2016年第3期133-138,共6页Journal of Forestry Engineering

基　　金：教育部高等学校博士学科点专项科研基金项目(20123204120019)

摘　　要：定制厨柜零件批量少,品种多,通常以订单为单位组织生产。笔者针对定制厨柜零件制造展开研究,优化定制厨柜生产模式,合并生产批量以提高定制厨柜制造企业的生产效率及材料利用率。通过采集定制厨柜零件结构及加工信息,采用模糊聚类算法对不同结构类型及工艺流程的零件进行分类,建立最优聚类零件族及分类制造模型并应用于企业生产。采用零件族分类制造模型后,企业生产效率得到提升,且在制品占用场地面积减少。模糊聚类算法可优化定制厨柜零件的分类,以零件族为基础的制造模型合并了加工批量,使定制厨柜生产符合大规模定制的要求,利于定制厨柜制造企业提升效益。Due to the small quantities but many varieties of the customized cabinet parts,the production and storage in factory are usually organized according to orders. As a result,the productivity is difficult to be increased and the storage cycle is long. Converting customized cabinet＇s manufacture mode into mixed orders can combine production batches and make parts as logistics unit instead of a single order,which is regarded as an effective way to improve the productivity and material utilization. In order to achieve this goal,this study uses fuzzy clustering algorithm to classify parts and create a manufacture model. The following steps are summarized： 1） collecting customized cabinet factory＇s manufacture basic data（ including products＇ data） and creating parts＇ production processes matrix; 2） classifying parts with different structures and processes and creating parts＇ manufacture similarity clusters using fuzzy clustering algorithm; 3） creating the optimal part family and part family classification of the manufacturing model based on parts＇ manufacture similarity clusters; 4） implementing part family classification manufacturing model to a customized cabinet producer,then collecting production data,such as production cycle per order,productivity per worker per hour. The application results showed that the productivity is increased,WIP（ working in process） area is decreased,and the production cycle is also decreased. It is implied that the fuzzy clustering algorithm is an effective way to classify parts with different structure and processes. The advantage is that the samples can be classified effectively when its separation property is not strong. The manufacture model based on parts＇ manufacture similarity clusters shows that it is an effective method to optimize customized cabinet parts＇ manufacturing. It changes order into part as factory＇s production and logistics unit. The similar parts are gathered and processed by the part family so that the waiting times in production ar

关 键 词：定制厨柜 模糊聚类算法 零件族 制造模型 生产效率 

分 类 号：TS657[轻工技术与工程]