基于知识驱动的混合式CAPP关键技术研究

发布时间：2018-03-21 05:39

本文选题：实例推理　切入点：知识驱动　出处：《中北大学》2012年硕士论文　论文类型：学位论文

【摘要】：尽管CAPP目前在应用中还不尽人意，但是其在制造业中的重要作用依然吸引了很多研究人员的兴趣。本文详述了CAPP的发展历史以及研究现状，展望了CAPP的发展趋势，分析了CAPP目前存在的一些问题。工艺设计是一个信息处理和转换的过程，从这个角度出发，本文从总体上构建了能进行高效工艺设计的混合式智能化CAPP系统的三层体系结构和实施流程，从而确定了构成系统的主要功能模块，该系统以人为主体，从知识驱动和实例推理两个角度来构建，并从以下三方面进行了研究：第一、知识的表达建模的关键技术。组织和构建了需求—能力工艺知识体系结构，该结构基于工艺过程的工艺需求关系，从零件工艺需求和加工环境能力角度出发，将工艺知识划分成零件信息、工艺信息、需求知识和能力知识。其中零件信息和工艺信息通过集成矩阵，从零件的特征层次与工艺的加工元层次进行集成构成了工艺实例。第二、基于实例推理的工艺设计。基于相似结构具有相似工艺的原理，使用空间维度的思想来构建工艺实例问题描述空间模型，用工艺实例问题描述与目标零件在该空间中的距离来表征工艺实例与目标零件之间的相似度。基于这种相似度评价和工艺实例的表示和组织结构，提出了一个基于四层分类结构的实例检索算法，并且在检索出合适的实例后，由一个基于匹配矩阵的工艺实例匹配和修改算法来实现工艺的实例推理设计。第三、基于知识驱动的创成式工艺设计。根据需求—能力响应的知识体系，从零件信息模型中提取出零件的工艺需求，采用加工元的形式在需求和生产环境的能力之间建立映射。然后使用遗传算法，以机床变换次数和刀具变换次数的加权组合最少作为适应度评价指标，对加工元进行排序并优化，最终生成优化的零件工序。根据上述研究，，开发了混合式CAPP原型系统，以某芯轴为例，生成了优化的加工工序，验证了本文提出的方法的可行性和有效性。
[Abstract]:Although the application of CAPP is not satisfactory at present, its important role in manufacturing industry still attracts the interest of many researchers. This paper describes the development history and research status of CAPP, and looks forward to the development trend of CAPP. Some problems existing in CAPP are analyzed. Process planning is a process of information processing and transformation. In this paper, the three-layer architecture and implementation flow of a hybrid intelligent CAPP system which can carry out efficient process planning are constructed, and the main function modules of the system are determined. It is constructed from two angles of knowledge driven and case-based reasoning, and is studied from the following three aspects:. Firstly, the key technology of knowledge representation and modeling is organized and constructed. The architecture is based on the process requirement relationship of the process process, from the point of view of part process requirement and machining environment capability. The process knowledge is divided into part information, process information, requirement knowledge and capability knowledge, in which part information and process information are integrated from the feature level of the part and the processing element level through the integration matrix to form a process example. Secondly, process planning based on case-based reasoning. Based on the principle that similar structures have similar processes, the spatial model of process case problem description is constructed by using the idea of spatial dimension. The similarity between the process case and the target part is represented by describing the distance between the process case and the target part in this space. Based on this similarity evaluation and the representation and organization structure of the process case, A case retrieval algorithm based on four-layer classification structure is proposed, and a matching matrix based process case matching and modifying algorithm is used to realize the case-based reasoning design of the process. Thirdly, generative process planning based on knowledge driven. According to the knowledge system of demand-ability response, the process requirements of parts are extracted from the part information model. In the form of machining element, the mapping between the requirement and the capability of the production environment is established. Then, using the genetic algorithm, the least weighted combination of the number of machine tool transformations and the number of tool transformations is used as the fitness evaluation index. Sort and optimize the machining elements, and finally generate the optimized parts working procedure. Based on the above research, a hybrid CAPP prototype system is developed. Taking a core shaft as an example, an optimized processing procedure is generated, which verifies the feasibility and effectiveness of the proposed method.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH162;TP391.72

【参考文献】