面向快速智能设计的产品实例数据精简与重构  

Case data reduction and reconstruction for rapid and intelligent product design

作　　者：王军[1] 张宗超 何平 李永欣[1] 刘福林 

机构地区：[1]燕山大学机械工程学院,河北秦皇岛066004

出　　处：《机械设计》2017年第11期22-26,共5页Journal of Machine Design

基　　金：国家自然科学基金青年科学基金资助项目(51405427)

摘　　要：基于实例的设计是实现快速设计和智能设计的重要方法。文中针对复杂产品进行相似性设计时存在的实例资源离散、数据量庞大和冗余数据多导致难以进行知识提取和利用的问题,提出了产品族基因的概念,建立了产品基因的描述方法。基于产品族基因对实例数据进行了精简,采用神经网络技术对离散基因数据进行了数据重构,重构后的数据结构为神经网络关系模型,该模型能够为产品智能快速设计提供有效的、统一的和结构化的数据支持,之后,建立了基于知识重构模型的快速设计系统结构,验证了模型的可行性。Case-based design is an important method for rapid and intelligent product design. Aiming at the problem that the case data is dispersed, huge and redundant, and it is often difficult for rapid design system to extract and utilize the case data, a product gene concept is presented and its description is discussed. Then the case data is reduced based on product family gene, case data structure is reconstructed based on artificial neural network, the reconstructed data structure is the neural network relation model. The model can support an efficient, common and structured dada structure for intelligent and rapid design system. Then, a prototype design system is presented and verified based on the model.

关 键 词：产品族基因 快速设计 知识重构 神经网络 

分 类 号：TH122[机械工程—机械设计及理论]