摘要
为充分利用产品概念设计中积累的成功经验和数据,提出基于粗糙集、支持向量机等理论的融合推理模型。利用模糊集理论对客户需求属性中的语义化、模糊化信息和连续值进行离散化处理,然后利用粗糙集理论对条件属性的冗余信息进行属性约简和规则提取,利用近邻算法获得产品设计的最相似实例。对于未找到相似实例的设计要求,利用支持向量机回归模型进行创新设计,通过人工调整参数,最终得到产品概念设计的最优方案。该模型建立了概念设计客户需求与产品质量特征之间的联系,克服了传统近邻算法的缺陷。基于UG平台开发出具有良好人机界面的采煤机概念设计原型系统,实现了该融合推理模型的工程应用,经实践验证,该方法较为客观、准确和高效。
To take full advantage of the successful experience and accumulated data of product conceptual design,fusion reasoning model was put forward,which was based on theories of fussy set,rough set and support vector machine.At first,by using fussy set theory,the discretization treatment of the semantization,fuzzy information and continuous value in customer demands attribute would be made.Then by utilizing rough set theory,the redundant information of condition attribute would be processed so as to simplify attribute and extract rules.After that,using nearest-neighbor algorithm,the most similar case design would be obtained.For instance not found similar design requirements,could use support vector machine regression model for innovative design.Finally,the optimal plan of product conceptual design could be available,as long as designers make artificial modifications for the design result.The model establishes the link between customer demands and features of product quality,overcomes the deficiencies in traditional nearestneighbor algorithm.At the same time,based on UG platform,the model could be used to develop conceptual design system of shearer with good man-machine interface.It is proved that the method is more objective,accurate and efficient.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2010年第10期1748-1753,共6页
Journal of China Coal Society
基金
国家"十一五"支撑计划重点课题资助项目(2007BAB13B01-02)
关键词
采煤机
概念设计
融合推理
粗糙集
支持向量机
shearer
conceptual design
fusion reasoning
rough set
support vector machine
