一种基于数据挖掘的制造业工厂设备布局方法

摘要

针对经典的求解单行直线型布局算法中需要大量参数、要求设备等概率使用的限制,提出了一种基于数据挖掘的制造业工厂设备布局方法FMDM.FMDM采用数据挖掘Apriori算法对已有的生产调度计划或柔性作业车间调度问题的调度解进行挖掘,根据贪心方法在频繁项的基础上获得的初步布局方案,给出了将候选方案进行筛选得到最终方案的算法CACULATE_EDIT_DISTANCE.实验结果表明:该方法可对无参数的初建车间进行有效的初步布局,不限制设备的使用概率,能实现多工件共享设备,多工件并发生产,且FMDM结果作为经典算法的输入可提高经典算法的收敛速度.%The algorithm for the design of a straight-line layout has been widely employed to solve facility layout problems . However , this algorithm requires sufficient parameters and demands equal probability of machines being used .This paper accordingly proposes an approach based on data mining to manufacturing facility layout , named as FMDM.FMDM firstly adopts the Apriori Algorithm for mining the existing production scheduling plan .Then a greedy algorithm is applied to the frequent item set, resulting in layout plans .Lastly, an algorithm is provided to select the best plan from these layout options.Experimental evidence shows that FMDM can be applied to a newly-built workshop′s facility layout planning without parameters and regardless of the probability of machines′use.This approach helps to achieve multiple jobs on the shared device and concurrent production .In addition, the rate of convergence can be increased through inputting the result of FMDM to traditional algorithms .