The facility layout problem belongs to typical multiple-attribute decision-making（MADM） problems. To make the information axiom fit to MADM problems, the original computation method for the information content is improved by increasing the satisfaction degree item. Attribute values are divided into precise type, uncertainty type and fuzzy type. For benefit, cost, fixation, and interval type attributes, the computation methods for the information content on the three types of attribute values are presented. The improved information content can reflect the system success probability and the decision-maker satisfaction degree simultaneously and evaluate the MADM problem including multiple type attribute values. Finally, as a case study, the facility layout alternatives of a welding assembly workshop are evaluated. The result verifies the validity and the feasibility of the improved information axiom on the MADM problems.
为提高模块变型设计的效率和准确性,提出了一种基于实例推理技术的模块化变型设计方法。将实例推理技术(case based reasonin,CBR)应用到模块化产品的变型设计中,CBR技术实现相似产品设计方案与相似结构模块的检索和选择,采用基于特征的实例框架表达方法,使用自组织映射神经网络,实现实例的可视化聚类,在此基础上构造分层多级实例库。采用基于差异和规则的实例修改方法,提高了CBR的检索速度和准确度。最后以卧式加工中心的变型设计为例,说明了该方法的可行性和实用性。
