基于加权指数平滑的里程仪故障检测方法  

Odometer Fault Detecting Strategy Based on Weighted Exponential Smoothing

作　　者：杨鹏翔[1,2] 秦永元[1] 朱启举[1,2] 

机构地区：[1]西北工业大学自动化学院,西安710072 [2]中国兵器工业第203研究所,西安710065

出　　处：《弹箭与制导学报》2010年第4期29-32,共4页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：国防科技重点预研项目资助

摘　　要：车载定位定向系统行进过程中滑行或打滑时，里程仪输出故障会显著降低系统精度。文中将惯性导航系统与里程仪的速度之差映射至有限区间，利用加权指数平滑的思想设计了检测时间窗随故障幅值自适应调整的里程仪故障在线检测方法。仿真结果表明，里程仪短时故障时，文中方法能够有效检测并及时告警，利用惯性导航系统速度隔离里程仪输出；故障消除后又能迅速解除故障警戒，恢复正常组合导航，提高了系统精度和可靠性。While a vehicle position and azimuth determining system（PADS） slipping or sliding in-flight, the odometer velocity output will remarkably debase the system navigation precision. In this paper, the error between inertial navigation system（INS） and odometer was projected to a finite interval, and then, a real-time odometer default detection strategy was designed through making use of the weighted exponential smoothing, which detection time window self-adjusting according to the fault magnitude. A mathematic simulation result show that, as a short-time odometer fault appears, this detection method can effectively detect and give alert to PADS rapidly, then isolate odometer output by INS velocity. When fault disappears, this method can de-alert and recover to normal navigation status, which decreases the error accumulation, improves the system precision and reliability.

关 键 词：定位定向 里程仪 故障检测 加权指数平滑 自适应 

分 类 号：V249.3[航空宇航科学与技术—飞行器设计]