摘要
为研究在定制家具分批生产之前的订单聚类问题,根据不同订单的信息,选取对定制家具排产影响较大的订单特征,将订单文本向量化,计算家具订单之间的相似度。利用蚁群算法的状态转移规则,通过聚类数目未知的蚁群算法将订单类别进行归并,以聚类完成后所有类别的材料种类总数最少为目标。通过仿真试验将其与聚类数目已知的蚁群聚类算法和K-means算法聚类结果进行比较,结果表明:相对于上述两种聚类算法,在定制家具订单聚类问题研究中,本文研究的算法聚类效果更好。
In order to study the order clustering problem before the batch production of customized furniture.In this paper,according to the information of different orders,the order features which have great influence on the production scheduling of customized furniture were selected,the order text was vectorized,and then the similarity between orders was calculated.Through using the state transition rule of ant colony algorithm,the order categories were merged by the ant colony algorithm with unknown clustering number.The goal is to minimize the total number of material categories in all categories after clustering.The simulation results were compared with those of ant colony clustering algorithm with known number of clusters and K-means algorithm.The results showed that compared with the above two clustering algorithms,the clustering effect of this algorithm is better in the research of customized furniture order clustering.
作者
陶涛
王洁
刘忠会
陈星艳
冯万福
TAO Tao;WANG Jie;LIU Zhong-hui;CHEN Xing-yan;FENG Wan-fu(Central South University of Forestry and Technology,Changsha 410004,China;Oupai Home Furnishing Group Co.,Ltd.,Guangzhou 510000,China;Jiangsu Kaidi Household Products Co.,Ltd.,Pizhou 221300,China)
出处
《林产工业》
北大核心
2020年第5期49-52,共4页
China Forest Products Industry
基金
湖南省教育厅科学研究重点项目(19A522)。
关键词
定制家具
订单
蚁群聚类算法
聚类分析
生产效率
Customized furniture
Orders
Ant colony clustering algorithm
Clustering analysis
The production efficiency
