[1]刘国燕,徐晓苏,白宇骏.基于H∞滤波算法的前向神经网络在SINS初始对准中的应用[J].东南大学学报(自然科学版),2003,33(3):331-334.[doi:10.3969/j.issn.1001-0505.2003.03.021]
 Liu Guoyan,Xu Xiaosu,Bai Yujun.Application of feedforward neural networks based on H∞ filter in SINS[J].Journal of Southeast University (Natural Science Edition),2003,33(3):331-334.[doi:10.3969/j.issn.1001-0505.2003.03.021]
点击复制

基于H滤波算法的前向神经网络在SINS初始对准中的应用()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
33
期数:
2003年第3期
页码:
331-334
栏目:
仪器科学与技术
出版日期:
2003-05-20

文章信息/Info

Title:
Application of feedforward neural networks based on H filter in SINS
作者:
刘国燕 徐晓苏 白宇骏
东南大学仪器科学与工程系,南京 210096
Author(s):
Liu Guoyan Xu Xiaosu Bai Yujun
Department of Instrument Science and Engineering, Southeast University, Nanjing 210096, China
关键词:
捷联惯性导航系统 非线性时变特性 前向神经网络 H滤波
Keywords:
SINS nonlinear time-varying characteristic feedforward neural networks H filter
分类号:
U666.12
DOI:
10.3969/j.issn.1001-0505.2003.03.021
摘要:
针对捷联惯性导航系统动基座初始对准的误差模型具有非线性、时变性的特点,提出将前向神经网络应用于该误差模型,并采用线性H鲁棒滤波算法在线调整网络权值,得到一种适合于该系统模型的改进的自适应前向神经网络.仿真结果表明,在存在模型误差或噪声不确定情况时,该网络不仅具有较好的鲁棒性能,而且能使误差在很短的时间内收敛,提高了系统的实时性,而对准精度与采用推广卡尔曼滤波器的精度相当.
Abstract:
According to the characteristic of nonlinear time-varying in strapdown inertial navigation system(SINS)initial alignment’s error model, the application of feedforward neural network to the error model is proposed by using linear H robust filter to adjust the network’s weights in real time. As a result an adaptive neural network is brought about for SINS. The simulation results show that when there exists model error or uncertain noises, the method has high convergence speed and better robustness comparing to the extended Kalman filter(EKF). In addition the precision is similar to EKF’s.

参考文献/References:

[1] Narendra K S,Parthasarathy K.Identification and dynamical systems using neural networks[J].IEEE Trans Neural Networks,1990,1(1):4-27.
[2] Simghal S,Wu L.Training feed-forward networks with the extended Kalman filter[A].In: Proceeding of IEEE International Conference on A Coustics Speech and Signal Processing[C].Giasgow,Scotland,1989.1187-1190.
[3] 顾成奎,王正欧.基于前向神经网络的非线性时变系统辨识[J].管理科学学报,2001,4(3):36-45.
  Gu Chengkui,Wang Zhengou.Nonlinear time-varying systems identification by feedforward neural networks [J]. Journal of Management Sciences in China,2001,4(3):36-45.(in Chinese)
[4] 马晓敏,周忙来.一种适于非线性系统辨识的神经网络学习算法[J].石油大学学报,1996,20(1):80-93.
  Ma Xiaomin,Zhou Manglai.Neural network learning algorithm suitable for nonlinear system identification [J]. Journal of the University of Petroleum,1996,20(1):80-93.(in Chinese)
[5] Jeff B Burl.H estimation for nonlinear systems[J]. IEEE Signal Processing Letters, 1998,5(8):199-202.
[6] Ball Joseph A,Kachroo Pushkin,Krener Arthur J.H tracking control for a class of nonlinear systems[J]. IEEE Transactions on Automatic Control,1999,44(6):1202-1206.
[7] Hassibi Babak,Sayed Ali H.Linear estimation in Krein spaces-part Ⅱ:applications [J]. IEEE Transactions on Automatic Control, 1996, 41(1):34-49.
[8] Nishiyama Kiyoshi,Suzuki Kiyohiko.H learning of layered neural networks [J].IEEE Transactions on Neural Networks, 2001,12(6):1265-1277.

备注/Memo

备注/Memo:
作者简介: 刘国燕(1977—),女,博士生,lgysd@163.com; 徐晓苏(联系人),男,教授,博士生导师,xxs@seu.edu.cn.
更新日期/Last Update: 2003-05-20