| 几种基于神经网络的导弹惯性器件故障预报方法及其性能比较 |
| |
| 引用本文： | 吕瑛洁,胡昌华.几种基于神经网络的导弹惯性器件故障预报方法及其性能比较[J].战术导弹技术,2005(2):52-55. |
| |
| 作者姓名： | 吕瑛洁 胡昌华 |
| |
| 作者单位： | 第二炮兵工程学院,西安710025 |
| |
| 摘 要： | 研究了标准BP网络、改进的BP网络(带动量的自适应BP网络)、 L-M网络和RBF网络及其学习算法, 探讨了基于这四种神经网络的导弹惯性器件故障预报方法, 并通过仿真实验对四种网络的预测预报性能进行了分析比较.结果表明, L-M网络和RBF网络对惯性器件的故障预报比两种BP网络更准确, 收敛速度更快.
|
| 关 键 词： | 故障预报 BP网络 L-M算法 RBF网络 |
| 文章编号： | 1009-1300(2005)02-0052-04 |
| 修稿时间： | 2004年9月22日 |
The Performance Analysis of Missile Inertia Device Fault Prediction Based on Neural Networks |
| |
| Lv Yingjie,Hu Changhua.The Performance Analysis of Missile Inertia Device Fault Prediction Based on Neural Networks[J].Tactical Missile Technology,2005(2):52-55. |
| |
| Authors: | Lv Yingjie Hu Changhua |
| |
| Abstract: | BP, fast BP, L-M and RBF neural network (NN) as well as their learn algorithms are studied. The prediction performances of these networks are analyzed and compared by studying on simulation experiment of missile inertia device fault prediction based on the networks. It is demonstrated that L-M NN and RBF NN have the quicker convergence speed, higher efficiency and better precision. |
| |
| Keywords: | fault prediction BP network Levenberg-Marquardt algorithm RBF (radial basis function) network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
