| 一种自组织神经网络及其应用 |
| |
| 引用本文： | 章新华,张晓明,林正青.一种自组织神经网络及其应用[J].火力与指挥控制,1997(1). |
| |
| 作者姓名： | 章新华 张晓明 林正青 |
| |
| 作者单位： | 海军大连舰艇学院，大连测控技术研究所 |
| |
| 基金项目： | 中国舰船研究院青年发展基金 |
| |
| 摘 要： | 提出了一种适用于目标识别和故障诊断的自组织神经网络（ＳＮＮ）模型和一种基于模糊关联度的判别准则。ＳＮＮ学习速度快，能在不破坏原有“知识”的情况下，学习、记忆新知识，且具有对外部环境的自适应能力。新的判别准则充分考虑了网络输出矢量各分量的大小因素及其与各期望目标矢量的相关程度，使类别判决更加合理可靠。本模型被用于被动声纳目标分类，取得了较好的效果。
|
| 关 键 词： | 自组织神经网络 目标识别 声纳目标 关联分析 |
An Adaptive Neural Network and its Application |
| |
| Zhang Xinhua,Zhang Xiaoming.An Adaptive Neural Network and its Application[J].Fire Control & Command Control,1997(1). |
| |
| Authors: | Zhang Xinhua Zhang Xiaoming |
| |
| Abstract: | An adaptive neural network suitable for target recognition and a fuzzy relational membership degree based classification criterion are proposed in this paper.The network which has rapid learning speed can learn new knowledge,meanwhile it doesn't desrupt the memorized knowledge.It has better adaptability to outer circumstance.The classification system based on new criterion,which considers more reasonable elements than conventional one,has better reliability.Satisfactory results could be obtained by using the model proposed here to classify passive sonar targets. |
| |
| Keywords: | self orgnization neural networks target recognition sonar target relational analysis |
| 本文献已被 CNKI 等数据库收录！ |
