基于实例和知识的夹具智能CAD系统研究

【作者】 康永刚；

【导师】 姜澄宇；

【作者基本信息】 西北工业大学 ， 机械制造及其自动化， 2004， 硕士

【摘要】 夹具作为一种重要的工艺装备在现代制造业柔性化发展趋势下，地位越来越重要。本文结合国家863计划重点项目“关中高新技术产业带网络化制造系统开发与应用”，开展了基于实例和知识混合推理的夹具设计技术研究，并完成了基于实例和知识混合推理的夹具设计系统的开发工作。该系统是关中高新技术产业带网络化制造系统的重要组成部分。 论文通过对夹具设计内容和特点的分析，提出了以CBR为核心、KBR为辅助混合推理的夹具设计新思路。由输入的设计信息出发，利用夹具设计领域知识进行KBR推理，并将符合设计要求的KBR推理结果作为CBR实例检索条件检索实例库得到相似的夹具设计实例，最后经基于CBR的实例修改模块修改以后，得到满意的夹具设计结果。 通过对夹具设计知识的表示方法、夹具设计的推理机制的研究，确定了多种知识表示方法相结合、正反向推理机制相结合的夹具方案设计思路。并实现了基于知识的夹具方案设计的自动选取。 通过对夹具实例的表示与信息模型的研究，提出了一个用于表示、检索及修改夹具设计实例的编码系统——FIXCOD。基于该编码系统，对夹具设计实例的检索与相似度计算进行了研究，提出了对关键因素的两重加权平均来定义相似系数的相似度计算算法。并在以上研究基础上建立了基于实例的夹具修改模型。 基于上述研究提出了一个用于夹具设计和修改的混合推理模型—HCKRM，研究了基于实例和知识的混合推理过程。基于此模型应用VB6.0、SQL Server2000在CATIA V5R8的基础上开发了一套实例和知识混合推理的夹具设计系统CKBFDS。更多还原

【Abstract】 As an important Tooling assembly, fixture shows its more and more important status with the development of modern manufacturing to the flexible direction. Combining with the key item of Advance Technology Development Project of China (863 Program) -"Networked Manufacturing System Developing and Applying in Advanced Technology Industry Area of GuanZhong", the paper has mainly researched hybrid CBR/KBR Reasoning Model of fixture design, and developed a Hybrid Reasoning of Case & Knowledge-Based Fixture Design System. The System is the important part of Networked Manufacturing in Advanced Technology Industry Area of GuanZhong.Based on the analysis of the contents and characters of fixture design, the author proposed a hybrid CBR/KBR fixture design method with CBR as its core, KBR as its assistant, and CBR supporting case-adaptation process. Using the design information, the designer can expect to obtain a satisfactory subset of possible configuration designs of a fixture from the KBR process. Then the designer, using this subset and other information, can expect to obtain a similar design of the fixture from the retrieving case database, finally use CBR adaptation technique to obtain the satisfied design of fixture.Based on the research on the representation of fixture design knowledge and the reasoning method of knowledge-based fixture design, we confirmed a fixture design method with multi-fixture design knowledge and hybrid reasoning methods. And achieved automated choose of the configuration design of fixture based on knowledge.Researched the representation of fixture design cases and the case information model, proposed a code system for the representation, retrieval and the adaptation of fixture design cases- FIXCOD, then, the paper researched the retrieval and the similarity calculation algorithm, finally proposed a CBR cases adaptation model.Based on these researches, the author proposed a hybrid CBR/KBR reasoning model - HCKRM. Based on it developed a Hybrid Reasoning of Case & Knowledge-Based Fixture Design System-CKBFDS with VB6.0 SQL Server2000 based on the CATIA V5R8 platform.更多还原