| ART1神经网络在基于CBR概念设计中的应用 |
| |
| 引用本文： | 薄瑞峰,黄洪钟,黄莉.ART1神经网络在基于CBR概念设计中的应用[J].测试技术学报,2005,19(4):447-452. |
| |
| 作者姓名： | 薄瑞峰 黄洪钟 黄莉 |
| |
| 作者单位： | 大连理工大学,精密与特种加工教育部重点实验室,辽宁,大连,116023;大连理工大学,精密与特种加工教育部重点实验室,辽宁,大连,116023;电子科技大学,机械电子工程学院,四川,成都,610054;中北大学,机械工程系,太原,030051 |
| |
| 基金项目： | 国家自然科学基金资助项目（59685003）;高等学校全国优秀博士学位论文作者专项基金资助项目（200232）. |
| |
| 摘 要： | 提出一种机构编码方法，可以分别对机构的运动功能属性和功能质量属性进行二进制编码，在此基础上，深入探讨了ART1神经网络在机构实例的分类和决策中的应用并对其算法进行了改进．与其它神经网络相比，ART1网络的优点在于可以实现机构动态分类和决策，设计者可根据自己的意图，通过调节警戒参数得到满意或最优的设计方案．实例表明，该方法合理可行．
|
| 关 键 词： | ART1网络 机构编码 动态分类决策 设计检索 CBR 机械运动方案设计 |
| 文章编号： | 1671-7449（2005）04-0447-06 |
| 收稿时间： | 2004-12-25 |
| 修稿时间： | 2004年12月25 |
Application of ART1 Neural Network in Conceptual Design Based on CBR |
| |
| B Rui-feng, HUANG Hong-zhong.Application of ART1 Neural Network in Conceptual Design Based on CBR[J].Journal of Test and Measurement Techol,2005,19(4):447-452. |
| |
| Authors: | B Rui-feng HUANG Hong-zhong |
| |
| Affiliation: | B0 Rui-feng, HUANG Hong-zhong |
| |
| Abstract: | A novel coding approach to mechanism is presented at first, by which the attributes of kinematic function and function quality for mechanism can be expressed respectively with a list of binary codes. Then based on an introduction to ART1 neural network, application of ART1 in classification and decision for mechanism cases is deeply investigated and some improvement for its learning algorithm is made. Compared with other neural network, the advantage of ART1 is that it can implement a dynamic classification and decision, viz. the designer can acquire an optimal design scheme or several satisfactory schemes in terms of his design intent by adjusting the value of vigilance parameter. |
| |
| Keywords: | ARTI mechanism coding dynamic classification and decision design retrieval CBR kinematic scheme design |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
