欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2015, Vol. 36 ›› Issue (11): 2190-2195.doi: 10.3969/j.issn.1000-1093.2015.11.026

• 研究简报 • 上一篇    下一篇

面向自行火炮变型设计问题的混合案例推理技术研究

羊柳1, 钱林方1, 丁晟春2, 尹强1   

  1. (1.南京理工大学 机械工程学院江苏 南京 210094; 2.南京理工大学 经济管理学院江苏 南京 210094)
  • 收稿日期:2015-03-06 修回日期:2015-03-06 上线日期:2016-02-02
  • 通讯作者: 羊柳 E-mail:ylnjust@126.com
  • 作者简介:羊柳(1989—), 男, 博士研究生
  • 基金资助:
    国防基础科研项目(A2620130003)

Research on Hybrid Case-based Reasoning Technique for Self-propelled Artillery Variant Design

YANG Liu1, QIAN Lin-fang1, DING Sheng-chun2, YIN Qiang1   

  1. (1.School of Mechanical Engineering,Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China;2School of Economics and Management, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China)
  • Received:2015-03-06 Revised:2015-03-06 Online:2016-02-02
  • Contact: YANG Liu E-mail:ylnjust@126.com

摘要: 自行火炮设计过程中存在大量变型设计问题,为提高自行火炮设计质量和效率,将案例推理技术引入自行火炮变型设计问题求解过程中。针对自行火炮案例特点及变型设计的需求,提出了一种混合案例推理技术。通过研究基于本体的自行火炮案例表示技术,构建了自行火炮设计案例本体,并确定自行火炮案例检索过程中不同情况下的相似度计算方法。在此基础上,采用基于神经网络的参数预测技术,实现了变型设计问题中设计参数的求解。并以复进机设计为例,验证了所提出方法的可行性和有效性。

关键词: 兵器科学与技术, 自行火炮, 变型设计, 案例推理, 本体, 神经网络

Abstract: In order to improve the design quality and efficiency of self-propelled artillery in consideration of a large number of variant design problems existing in self-propelled artillery design, the case-based reasoning technology is introduced into the solution procedure for self-propelled artillery variant design problem. The process of hybrid case reasoning is determined according to the characteristics of self-propelled artillery design case and the requirements of variant design. The ontology-based case representation technique is studied, and a self-propelled artillery design case ontology is built.The similarity calculation formulas for different scenarios are determined for the self-propelled artillery case retrieval. On this basis, the solution of design parameters for variant design problem is achieved by using neural network-based parameter prediction technique. The feasibility and effectiveness of the proposed method are demonstrated by taking recuperator design for example.

Key words: ordnance science and technology, self-propelled artillery, variant design, case-based reasoning, ontology, neural network

中图分类号: