以装拆时间、装拆难度、材料成本、制造成本、再设计和后处理成本,以及产品质量等的变化程度作为零件是否合并的条件属性,零件是否合并作为决策属性,应用粗糙集理论分析设计过程中的专家经验和知识。在经典粗集理论的基础上,以属性重要度作为启发式信息来计算属性核和属性约简,并以此挖掘隐含在经验数据中的专家知识,形成零件是否可以合并的决策规则。最后,以小型塑封机再优化设计为例,验证了该方法的实用性和有效性。
A methodology based on Rough Set (RS) was developed to analyze expertise and acquire expert knowledge wrapped in experienced data. It regarded assembly/disassembly time, assembly/disassembly difficulty, material cost, manufacturing cost, redesign cost and quality cost as condition attributes, while taking component integration as decision attribute. Based on classical RS, attribute importance was used to calculate the core and reduction of attribute as heuristic information, and then expertise wrapped in experienced data was mined. An instance was provided to illustrate the feasibility and effectiveness of the proposed method.