基于遗传算法的产品结构聚类重组

通过遗传算法实现了产品结构的智能聚类重组,把绝大部分元素之间的联系都包含在了各聚类内部以及bus类中,大大降低了产品结构各聚类之间的耦合强度。在算法的建模过程中,给出了产品结构DSM模型二维编码方法,并在此基础上给出了基于联系权重的适应度函数和选择概率的求解算法。最后通过对一个简单实例产品结构DSM模型的智能化聚类重组,验证了算法的可行性。

Clustering and Reorganization of Product Architectures Based on Genetic Algorithm

The intelligent clustering and reorganization of product architectures are carried out by using genetic algorithm.With the relations among the majority parts of elements encompassed within clusters and bus clusters,the coupling strength among clusters is reduced to a great extent.In the process of modeling the algorithm,a two-dimensional encoding method for the design structure matrix(DSM) model of product architectures was presented.On the basis of the method,the solution algorithms for fitness functions and selectional probability based on relational weights were provided.Finally the intelligent clurtering and reorganization of a simple instance of product architecture in the DSM model verify the feasibility of the algorithm.