【作者】 王珂；

【导师】 陆一平；

【作者基本信息】 北京交通大学 ， 机械工程， 2012， 硕士

【摘要】 现代设计理论认为,产品的结构、功能和成本的70%在设计阶段就被决定了。材料的选择又是机械产品设计工作中一个极其重要的组成部分。很多机械的重大失效并不是在于它的运动设计和动力设计,而在于材料的种类和牌号选择不当。近年来,越来越多的新材料被使用,以减轻产品的重量,提高产品的质量,降低产品的成本。另外,创新设计也需要材料选用方面的创新给予支持。因此如何在数量巨大的材料库中选择最优的材料就成了国内外学者研究的热点。本文研究和初步实现了一个智能材料选择系统,建立了材料数据库和知识库,着重研究了BP神经网络和专家系统在材料选择方面的应用。本文的工作包括以下几个方面：一、论述了传统材料选择过程中存在的问题,分析了国内外关于选材方法和专家系统的研究现状。分析了神经网络与专家系统各自的优缺点后,本文提出以BP神经网络和专家系统相结合的方式设计选材系统,为设计人员推荐材料。二、确定了材料数据来源,研究了材料的标准存储格式后,建立了包含大多数工程材料信息的材料数据库,并开发了系统的材料查询模块。建立知识库,系统通过学习样本和总结归纳两种方式,获取专家的选材知识并保存在知识库中。三、根据知识库中的知识建立系统的选材模块,包含一个BP神经网络和一个决策树,通过提问的方式为设计人员推荐材料,并且能够对选材结果做出合理的解释。四、建立知识管理模块、用户反馈模块、帮助模块、用户管理模块等辅助模块,完善系统功能。系统采用B/S结构,所有服务都在服务器上实现,减少客户端载荷,使用户能够更加简单方便的使用。五、文章最后通过一个应用实例,验证了系统的有效性。本系统不仅可以给设计人员推荐合适的材料,也可以为经验不足的用户积累选材知识,具有广泛的应用价值。更多还原

【Abstract】 The Modern Design Theory holds that decisions made during the design period determine70%of the product’s structure, function and cost. Material selection is an extremely important part of the design of mechanical products. In recent years, many new materials are adopted to reduce weight, improve performance and reduce the costs. A lot of major mechanical failure is not because of the kinematic design and dynamic design, but because of the inappropriate choice of material types and grades. With the emergence of a large number of new materials, the innovative design material selection aspects of innovation to support. Therefore, how to choose the best materials in a huge number of materials has become the research hotspot.This paper research and develop an intelligent material selection system; establish a materials database and a knowledge base and focused on the application of BP neural network and expert system for material selection. This work includes the following aspects:1. We discussed the problems in the traditional method of material selection and analyzed the status of research on material selection methods and expert systems of domestic and international and introduced the theoretical knowledge of neural networks and expert systems. After analyzing the advantages and disadvantages of neural networks and expert systems, we developed a material selection system applying a material-selection strategy combined BP neural network and expert system to select the suitable material for designer.2. Found material data sources; Researched the standard storage format of the material; established a materials database that contains the information of most engineering materials and developed the material query module. We established a knowledge base. The system obtains the material selection knowledge of experts through learning sample and summarizing and stored the knowledge in the knowledge base.3. Developed the material selection module with the knowledge stored in the knowledge base, which contains a gear selection BP neural network and a decision tree. The system selects the suitable material for designer with the form of questions and be able to make the reasonable explanation of the results.4. Developed the knowledge management module, the user feedback module, the help module and the user management module to improve the function of the system. The system adopts B/S structure and all services on the server in order to reduced the load of client. Users can use the system easily and comfortably.5. The system’s effectiveness was demonstrated by a material-selection case in actual applications. This system could help designer select suitable materials and provide knowledge for inexperienced engineer. So it has great application significance.更多还原