【作者】 蒋维；

【导师】 王成恩； 竺长安；

【作者基本信息】 中国科学技术大学 ， 精密仪器及机械， 2008， 博士

【摘要】 随着市场需求的分化和买方市场的形成,市场竞争中消费者需求呈现多样化、差异化和个性化。部分领域的客户已经开始要求能够完全定制产品.如何利用快速设计、优化来解决这一类个性化产品定制问题,降低产品成本,提高企业响应速度成为众多学者关注的问题,也是本文的主要研究内容。论文首先分析了目前锻压机床企业面临的问题,结合国内外快速设计、优化方法的研究情况,指出了目前锻压机床设计、优化中的一些不足之处,最后提出了论文解决的关键问题和研究意义。针对目前锻压机床设计过程与优化过程集成度不高的问题,本文尝试建立了能同时兼容设计与优化这两个环节所需要的锻压机床CAD/CAE混合模板库。将快速高效的参数化方法从一般的模型设计深入到有限元结构仿真分析,以实现在调整设计参数之后能够迅速得到新的设计模型,并迅速得到对新的设计模型的分析结果。在此基础上,提出了基于混合模板库的锻压机床产品快速设计、优化的技术体系结构.随后,本文结合基于特征的有限元模型的抽象与简化原理,并基于Pro/Engineer阐述了从CAD模型到CAE模型建立过程.以仿蜂窝结构的锻压机床横梁为例,论述了利用APDL建立了其参数化有限元分析模型的方法。然后使用XML语言描述了基于APDL的参数化有限元分析模型,并采用XSD定义了其前处理,后处理的过程,为结合客户需求快速建立有限元分析模型提供了基础。作为混合模型库的锻压机床产品快速设计、优化方法的一种具体实现,本文提出了一种新的产品设计模式-客户需求驱动的锻压机床产品快速设计系统。该系统从客户需求出发,得到产品设计需求目标,并采用设计结构矩阵(DSM)来进行设计活动的协调安排。然后通过模板匹配,提取混合模型库中的CAD模板信息,再根据设计知识快速求解设计参数。对于这种新型设计模式采用到的各项技术,结合Pro/Engineer平台的强大三维建模功能,利用Pro/Toolkit开发工具在Pro/Engineer中给出了系统的实现,提供了成功的具体应用实例。然后本文阐述了利用混合模板库,结合客户需求快速建立有限元分析模型。为此首先介绍了产品CAE有限元模型的快速建立过程中的数据流向、有限元分析特征的确定以及从混合模板库中CAE模板信息的提取方法。其次介绍了基于XML的通用数据交换技术,CAD/CAE模型数据XML表达,最后介绍了锻压机床典型产品CAD/CAE模型快速设计、优化系统实现。应用前文建立的有限元分析模板,本文尝试将Kagome和Triangle蜂窝结构应用于液压机横梁,进行结构创新性设计。其中针对由于Kagome结构突变导致优化过程不收敛的情况,采用动态搜索轨迹的混合遗传算法实现结构参数的随机优化,取得了较好的实际效果。本课题得到国家高技术研究发展计划(863计划)资助(No.2007AA04Z1A2),是其研究内容的一部分。更多还原

【Abstract】 Along with the division of market demand and the buyer’s market coming into being, the consumers’ demands such as multiplicity,diversity and individuation are getting considerably important.Some consumers started to customize the product.Therefore, how to improve the customized producing through distinctive rapid response design while reduce cost and response time has become a key emphasis.Firstly in the present article,the author analyzes the problems the forging machine enterprises facing,and points out the deficiency of the present design and optimization. Then the author proposes the key problem to solve’and the significance of the research.According to the problem of unconsolidated design and optimization process,the author tries to establish a compound CAD/CAE forging machine parts templates library, in order to apply the efficient parameter methods to both CAD model design and finite element analysis(FEA).By that,the regenerated CAD model and the CAE analysis result can be obtained rightly after parameters regulation.Finally,the author proposes a compound templates library based rapid design and optimization method architecture,which includes both CAD and CAE templates of forging machine parts.Subsequently,the paper presents the abstraction and abbreviation principle of feature based FEA model,and formulates the process from CAD model to CAE model by Pro/Engineer system.Then taking honeycomb-like beam as example,discusses the establishment of APDL parameter FEA model.Finally,the author uses XML to describe a APDL based FEA template,and defines the preparation and posttreatment process by XSD,by that,we establish a solid foundation to fast create FEA model combining custom requirements.As a case of compound CAD/CAE forging machine parts templates based rapid design and optimization method,the author proposes a new product design type-custom requirements driven forging machine rapid design system.The system starts from custom requirements to obtain the design goal,and applies DSM to coordinate the design tasks.By templates matching,the author extracts the CAD template from compound templates library,then obtains the design parameter by design knowledge.Based on Pro/Engineer platform,a detailed solution of several key technologies used in this system is provided.At the end,a successful application of this new design system is introduced.Following the compound templates library,the author sets up a FEA model generation method merged with custom requirements.Firstly,the paper sequentially formulates the dataflow of the modeling process,the determination of FEA feature and the method of extracting templates.Then,the author introduces a kind of data exchange method between CAD/CAE system and the representation of CAD/CAE model by XML.Finally,the paper sets forth the realization of forging machine rapid design and optimization system.At last,the author attempts tóapply the honeycomb-like structure to forging machine’s beam structure optimal design.To solve the problem of optimization divergence, the author uses a kind of hybrid genetic algorithm based on dynamic searching method to stochastic optimization,and the result indicates the efficiency of the algorithm.As a part,this task is sponsored by the Chinese National Programs for High Technology Research and Development(No.2007AA04Z1A2).更多还原