中国惯性技术学报
中國慣性技術學報
중국관성기술학보
JOURNAL OF CHINESE INERTIAL TECHNOLOGY
2013年
3期
406-410
,共5页
翁浚%成研%秦永元%严恭敏
翁浚%成研%秦永元%嚴恭敏
옹준%성연%진영원%엄공민
捷联惯导系统%故障检测技术%残差卡方检验%车辆运动约束
捷聯慣導繫統%故障檢測技術%殘差卡方檢驗%車輛運動約束
첩련관도계통%고장검측기술%잔차잡방검험%차량운동약속
strapdown inertial navigation system%FDI%error 2? detection%vehicle constraints
为了提高捷联惯导系统与里程计组合导航系统在里程计发生故障后的导航性能,在传统的故障检测算法的基础上,考虑了车辆运动约束条件在陆地导航系统中的特点,设计了两层基于残差的卡方故障检测算法,推导了在里程计故障后的车辆约束量测构造与滤波方案。采用传统单层与两层故障检测算法对80 km 的车载导航数据进行了处理,试验证明,传统单层算法在里程计出现时长为2000 s 的输出故障后,定位无法再次收敛。两层故障检测算法在里程计出现故障后能够顺利完成导航任务,里程计故障期间位置导航精度与无故障时相比仅相差2 m,提高了里程计发生故障后的导航性能。
為瞭提高捷聯慣導繫統與裏程計組閤導航繫統在裏程計髮生故障後的導航性能,在傳統的故障檢測算法的基礎上,攷慮瞭車輛運動約束條件在陸地導航繫統中的特點,設計瞭兩層基于殘差的卡方故障檢測算法,推導瞭在裏程計故障後的車輛約束量測構造與濾波方案。採用傳統單層與兩層故障檢測算法對80 km 的車載導航數據進行瞭處理,試驗證明,傳統單層算法在裏程計齣現時長為2000 s 的輸齣故障後,定位無法再次收斂。兩層故障檢測算法在裏程計齣現故障後能夠順利完成導航任務,裏程計故障期間位置導航精度與無故障時相比僅相差2 m,提高瞭裏程計髮生故障後的導航性能。
위료제고첩련관도계통여리정계조합도항계통재리정계발생고장후적도항성능,재전통적고장검측산법적기출상,고필료차량운동약속조건재륙지도항계통중적특점,설계료량층기우잔차적잡방고장검측산법,추도료재리정계고장후적차량약속량측구조여려파방안。채용전통단층여량층고장검측산법대80 km 적차재도항수거진행료처리,시험증명,전통단층산법재리정계출현시장위2000 s 적수출고장후,정위무법재차수렴。량층고장검측산법재리정계출현고장후능구순리완성도항임무,리정계고장기간위치도항정도여무고장시상비부상차2 m,제고료리정계발생고장후적도항성능。
For the sake of improving the navigation performance of strapdown inertial navigation system and odometer integrated system when odometer breaks down, a two-layer residual error 2? detection algorithm was designed based on traditional fault detection algorithm and by considering the constraints of vehicle motion. The construction and filtering scheme of vehicle constraints’s observation after odometer breaks down were deduced. An 80 km vehicle navigation test’s data was treated with traditional algorithm and two-layer detection algorithm respectively. The results verify that the positioning error of single-layer algorithm will not converge after odometer breaks down for 2000 s, and the two layer fault detection algorithm can succeed in the whole navigation task. During the odometer breaks down, the positioning error is only 2 m worse than that under no failure condition. When with break down odometer, the navigation performance by 2-layer fault detection algorithm is significantly higher than that by traditional algorithm.