Abstract: In order to meet the multi-variety and small-batch markets' demands and dynamic fluctuating customers' demands under mass customization and industrial cluster, and to meet the low cost and high robustness layout requirements of workshops,a multi objective layout optimization model was designed which included unit area layout cost, unit product logistics cost and layout entropy. A clustering parallel multi-objective genetic algorithm was proposed based on Pareto optimization. Fuzzy C-means clustering algorithm was introduced to improve the diversity and uniformity of Pareto solution set distributions. The multiple cell differential evolution re-insertion operation and the migration operation based on “elite strategy” were designed to enhance the global and local search ability of the algorithm, and effectively avoid the phenomenon of precocity. By comparing the typical examples, the validity of the algorithm and model was verified. In addition, the model may provide high practicability in real layout applications since it may reach the aim of low cost, and control the entropy of workshops within an ideal range.

Key words: polymorphic workshop, layout entropy, low entropy robust optimization, Pareto solution, fuzzy mean clustering

摘要： 为满足大规模定制和产业集群下多品种、小批量的市场需求和订单动态波动的客户需求以及车间低成本、高稳健性的布局要求，设计了以单位面积布置成本、单位产品物流成本和布局熵为优化指标的多目标布局优化模型。提出了基于Pareto优化的聚类并行多目标遗传算法，引入模糊C-均值聚类算法以提高Pareto解集分布的多样性与均匀性，设计了多元胞差分进化重插入操作与基于“精英策略”的移民操作，增强了算法全局与局部搜索能力，有效避免了早熟现象。通过典型算例对比，验证了模型和算法的有效性；同时在企业布局实例应用中，获得了既能满足低成本又能将布局熵值控制在理想范围内的车间布局方案，表明模型具有良好的实用性。

关键词: 多态性作业车间, 布局熵, 低熵稳健优化, Pareto解, 模糊均值聚类