成衣制造工艺知识可拓表示与实例推理应用研究  

Research on Extension Representation of Garment Manufacturing Process Knowledge and Case-based Reasoning Application

作　　者：祝勇仁[1] ZHU Yong-ren(Institute of Intelligent Manufacturing,Zhejiang Institute of Mechanical&Electrical Engineering,Hangzhou 310053,China)

机构地区：[1]浙江机电职业技术学院智能制造学院,杭州310053

出　　处：《北京服装学院学报（自然科学版）》2021年第2期56-63,共8页Journal of Beijing Institute of Fashion Technology：Natural Science Edition

基　　金：浙江省科技厅公益技术研究计划项目(LGG21E050012);校级科教融合重大研发孵化项目(A-0271-20-106)。

摘　　要：激烈的市场竞争,使企业必须满足客户需求进行多品种、小批量生产,导致企业的工艺设计方案急速增长。现有CAPP系统知识化表示不足,使工艺设计人员难以快速找到相似工艺,降低了工艺设计效率。通过对传统的谓词逻辑、产生式规则、语义网络和框架等知识表达方法进行可拓表示,对成衣制造工艺知识进行建模与实例推理应用研究,建立了定性与定量相结合的成衣制造可拓知识表达模型,提出了可拓信息物元相关性和相似性计算方法。把可拓学与实例推理相结合,构建了可拓激励推理模型,并将可拓学中的可拓距运用于该算法,进行实例搜索,根据距值的不同确定实例点在目标区间的位置。将激励算法与物元的可拓性和可拓距相结合,能提高激励算法的经验性,只要实例库中存在满足求解的相似工艺方案,则通过大量的应用和系统的自学习功能,可以拓展实例库,大大提高工艺设计效率,从而进一步增强CAPP系统的智能性及实用性。In order to adapt to the fierce market competition, enterprises must meet the needs of customers for many varieties, small batch production, which lead the incremental growth of enterprise process design. Insufficient knowledge representation of the existing CAPP system makes it difficult for technology design personnel to quickly find similar process and reduces the efficiency of process design. By using extension method to express traditional knowledge such as predicate logic, production rule, semantic network and frame, this paper modeled the process knowledge of garment manufacturing and studied the case reasoning application. It established qualitative and quantitative knowledge representation model based on extension, and put forward method of the matter-element extension information relevance and similarity calculation. It constructed the model of extension stimulant reasoning by combining extenics with case-based reasoning. It also applied the extension distance to this algorithm for case searching and determined the position of the instance in the target range according to the distance value. The combination of encourage algorithm with the matter-element extension and extension distance can improve the experiential of algorithm. As long as there are process design instances that can solve the similar process in the library and case database can be expanded with a large number of applications and the self-learning function of the system, the efficiency of process design can be greatly improved and the intelligence and practicability of CAPP system can be further enhanced.

关 键 词：成衣制造工艺知识 可拓 实例推理 可拓激励推理 可拓距 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]