摘要
在联邦滤波器结构的基础上 ,设计了一种基于神经网络的天文惯性导航系统 /全球卫星定位系统 /战术导航系统 (SINS/GPS/TAN)容错组合导航系统。该系统首先在各子滤波器中用神经网络滤波 (NNF)代替传统的卡尔曼滤波 ,然后用模糊神经网络 (FNN)对各子系统进行故障检测 ,最后由神经元进行全局融合。通过仿真 。
A stellar inertial navigation system/global position system/tactical navigation system fault tolerant integrated navigation system based on neural network is designed in this paper. The traditional Kalman filter is replaced by the neural network filter in each sub filter. Then the fuzzy neural network is used for detecting fault. The neurons are used for information fusion at last. The simulation results show that this approach is available.
出处
《上海航天》
2003年第2期22-26,共5页
Aerospace Shanghai
关键词
神经网络
天文惯性导航系统
全球卫星定位系统
战术导航系统
容错组合导航
故障检测
Stellar inertial navigation system
Globe position system
Tactical navigation system
Neural network
Integrated navigation
Fault detection
