神经网络信息融合方法在分布式捷联基准中的应用

Information Fusion Based on Neural Networks in Distributed Reference

针对舰船所处的恶劣环境 ,以及舰船导航和武器系统对精度的要求越来越高和分布式指控系统的发展 ,提出了由安装在舰船甲板上有限个捷联基准、舰载平台罗经和利用测量结果推算出的甲板挠曲姿态信息构成分布式捷联基准系统 (简称分布式基准 )。然后基于径向基函数神经网络的信息融合技术建立全舰统一姿态基准系统。最后通过对简化的甲板模型的实验 。

In view of the adverse environment of a ship, the high accuracy requirement for the attitude determination in the marine navigation and weapon system, and the development of the distributed command system, a distributed strapdown attitude reference system (DSARS) is built. It is composed of some strapdown references on the deck of ship, the stabilized gyrocompass, and the deck deflection attitude information computed from measurement results. Then the unite attitude reference system of the whole ship is built based on the information fusion on the radial basis function neural networks. Finally, experiment on a simple model shows that the method is effective and has a higher precision.