| 一种基于多层SOFM网络的星图识别方法 |
| |
| 引用本文： | 何爱香,朱云华,安凯. 一种基于多层SOFM网络的星图识别方法[J]. 上海航天, 2007, 24(2): 25-29 |
| |
| 作者姓名： | 何爱香 朱云华 安凯 |
| |
| 作者单位： | 1. 山东工商学院,信电学院,山东,烟台,264005
2. 山东航天电子技术研究所,山东,烟台,264000 |
| |
| 摘 要： | 提出了一种基于自组织特征映射(SOFM)神经网络的星图识别算法。用基于支持向量机(SVM)的动态阈值选取算法选取导航星构建导航星库,将一多层多个并联SOFM子网的识别系统用于星图识别。给出了方法的流程。仿真结果表明,SOFM网络可提取星图中的复杂特征识别导航星。与传统三角形算法相比,该识别算法的识别准确率、鲁棒性和实时性更优,有一定的实用价值。
|
| 关 键 词： | 星图识别 自组织特征映射网络 三角形算法 输入样本 层次结构 |
| 文章编号： | 1006-1630（2007）02-0025-05 |
| 修稿时间： | 2006-03-07 |
Application Research on SOFM Networks in Star Pattern Recognition |
| |
| HE Ai-xiang,ZHU Yun-hua,AN Kai. Application Research on SOFM Networks in Star Pattern Recognition[J]. Aerospace Shanghai, 2007, 24(2): 25-29 |
| |
| Authors: | HE Ai-xiang ZHU Yun-hua AN Kai |
| |
| Affiliation: | 1, School of Information and Electronics Engineering, Shandong Institute of Business and Technology, Yantai Shandong 264005, China; 2, Shandong Aerospace Electronic-Technology Institute, Yantai Shandong 264000, China |
| |
| Abstract: | A novel method using self-organizing feature maps(SOFM) networks was presented for autonomous star pattern recognition in this paper.The guide star catalog was created using the support vector machine based automatic selection algorithm. And an identification system consisting of several SOFM hierarchical networks was designed and input patterns from the star distribution around each guide star were employed to train the multi-layer SOFM networks.The flowchart of the method was presented.The simulation results showed that the new method had better robustness to noise,real-time and accuracy of identification than the conventional triangle algorithm,which would be valuable in engineering. |
| |
| Keywords: | Star pattern recognition Self-organizing feature maps network Triangle algorithm Input pattern Layer structure |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
