摘要
针对传统残差χ^(2)检测方法在组合导航系统中渐变故障检测率不高的问题,为了及时有效地检测故障信息,提高系统可靠性,提出了一种基于AR量测建模的组合导航系统渐变故障双阈值检测方法。该方法通过建立无故障条件下量测数据的AR模型,结合卡尔曼滤波模型得到量测预报值进行残差计算,提高故障检测的灵敏度;搭建双阈值检测门限,对误警率和漏警率之间的受污染的量测数据,采用双阈值门限进行分类处理,降低了漏警率对数据可靠性的影响。避免由于传统残差χ^(2)检测方法因引入观测污染数据,对渐变故障不敏感的问题。为了验证所提方法的有效性,将该方法应用到SINS/GNSS组合导航系统中进行仿真实验。仿真结果表明,所提方法渐变故障检测漏警率较常规残差χ^(2)双阈值检测方法降低69%以上,整体滤波精度提高19%以上,提高了系统的可靠性。
Aiming at the problem that the traditional residual chi-square detection method is not high in the detection rate of gradual faults in the integrated navigation system,a double threshold gradual fault detection method based on AR measurement modeling is proposed to detect the fault information timely and effectively and improve the reliability of the system.In the method,the AR model of the measured data is established under the no-fault condition,and the residual calculation of the measured prediction value is obtained by the Kalman filter model,so as to improve the sensitivity of fault detection.At the same time,a double threshold detection threshold is set up,and the contaminated observation datas between the false alarm rate and the missing alarm rate are classified and processed by the double threshold,which reduces the influence of the missing alarm rate on data reliability and avoids the problem that the traditional residual chi-square detection method is insensitive to gradual fault due to the introduction of observation pollution data.In order to verify the effectiveness of the proposed method,the method is applied to the SINS/GNSS integrated navigation system for simulation experiments.The simulation results show that the detection missing alarm rate of the proposed method is reduced by more than 69%,and the overall filtering accuracy can be increased by more than 19%,which improves the reliability of the system.
作者
吕旭
胡柏青
戴永彬
高端阳
LYU Xu;HU Baiqing;DAI Yongbin;GAO Duanyang(College of Electrical,Naval University of Engineering,Wuhan 430033,China;School of Electrical Engineering,Liaoning University of Technology,Jinzhou 121001,China)
出处
《中国惯性技术学报》
EI
CSCD
北大核心
2021年第1期133-140,共8页
Journal of Chinese Inertial Technology
基金
国家自然科学基金(61703419,61873275)。
关键词
组合导航
残差
AR模型
故障检测
integrated navigation
residual
AR model
fault detection