基于聚类算法的虚拟单元布局研究

A Study about Virtual Cell Layout Based on Clustering Algorithms

以制造单元为研究对象,建立了以聚类算法为基础的虚拟单元布局数学模型。模型在考虑设备放置位置惟一性的基础上,以相邻设备间关联度总和最大为目标函数。通过Jaccard相似系数求得设备间关联度后,采用遗传算法进行模型的求解,并对染色体编码和遗传算子进行了改进。最后,通过算例分析证明所提出的方法能够很好地解决可重构系统中的虚拟单元布局问题。

This article focuses on the manufacturing cellular, and establishes a mathematical model about virtual cell layout based on clustering algorithms. According to the uniqueness of facility positions, this model takes the maximize sum total correlation degree of adjacent equipment as the objective function. It firstly obtains the correlation degree of adjacent equipment by adopting Jaccard similarity coefficients. The objective solution is then finished with a modified genetic algorithm, where both the chromosome coding and genetic operators are improved. Finally, an example validation indicates that this technology can well solve the problem of virtual cell layout of reconfigurable manufacturing system.