一种高超声速飞行器组合导航故障检测方法  

A Way of Integrated Navigation Fault Detection of Near Space Hypersonic Cruising Aircraft

作　　者：李海林[1] 张斌[1] 吴德伟[1] 卢虎[1] 

机构地区：[1]空军工程大学信息与导航学院,陕西西安710077

出　　处：《系统仿真学报》2017年第8期1809-1814,共6页Journal of System Simulation

基　　金：国家自然科学基金(61603412);陕西省自然科学基础研究计划(2017JQ6027)

摘　　要：飞行器高超声速飞行时,多普勒频移增大,GNSS捕获跟踪易丢失,CNS也会造成光线传播误差,使天体跟踪出现问题。提出了一种高超声速飞行器残差χ2-Fuzzy ARTMAP快速神经网络INS/GNSS/CNS组合导航故障检测方法,给出了高超声速飞行器INS/GNSS/CNS组合导航系统故障诊断原理,推导了残差χ2检验法的检测函数公式和Fuzzy ARTMAP快速神经网络原理算法,研究了残差χ2-Fuzzy ARTMAP快速神经网络组合导航故障检测实现方法。仿真结果表明该检测方法比传统残差χ2检验法更准确及时有效实现故障检测和隔离,并且得到的姿态、位置和速度信息精度更高,提高了临近空间高超声速飞行器INS/GNSS/CNS组合导航体系的精度和可靠性。The doppler shift is augmented, that causes acquisition and tracking of GNSS losing, error of CNS＇s ray propagation also causes the problem of celestial body tracking when aircraft is hypersonically flighting. A way of INS/GNSS/CNS integrated navigation fault detection of the hypersonic cruising aircraft based on the residual chi-square-Fuzzy ARTMAP （Adaptive Resonance Theory Map） fast neural networks was proposed. The fault diagnostic elements of INS/GNSS/CNS integrated navigation system of the hypersonic cruising aircraft were given; the detection function formula of residual chi-square test and Fuzzy ARTMAP fast neural networks arithmetic was deduced, the realizing way was studied. The results demonstrate that the test way is available and feasible, and the navigation precision and reliability are improved.

关 键 词：临近空间高超声速飞行器 组合导航 故障检测 残差X2检验法 FuzzyARTMAP神经网络 

分 类 号：V448[航空宇航科学与技术—飞行器设计]