基于无人机运动模型的双余度惯导故障诊断*

Double Redundancy Inertial Navigation Fault Diagnosis Based on UAV Motion Model

摘要:

捷联惯性导航设备作为无人机核心设备之一,其故障检测和隔离直接关系到无人机的飞行安全。目前对无人机系统故障诊断研究成果较多,但在实际工程应用中仍存在较大难度；针对双余度惯导系统难以通过互相对比的方法隔离缓慢漂移故障的难题,提出了基于无人机运动模型的双余度惯导故障诊断方法,通过无人机的遥测数据和气动参数对无人机进行局部建模,引入一种差值累计对比算法,搭建了一种用于双余度惯导的故障诊断系统,此系统内植于地面专家辅助软件可实现惯导缓慢漂移故障的诊断和隔离,另外,结合基于故障树的专家诊断系统结果,可在地面专家辅助软件中实时监控双惯导的故障状态；当某一惯导发生故障时,地面专家辅助软件可实时报警,通过显著提示辅助操作人员速准确地定位和确定故障种类并进行最终决策；通过故障数据回放试验验证了所提出方法的可行性与有效性。

Abstract:

Strapdown inertial navigation system (SINS) is one of the core equipment of unmanned aerial vehicle (UAV), it’s fault detection and isolation are directly related to the flight safety of UAV. At present, there are many research results on the fault diagnosis of UAV system, but there are still significant difficulties in practical engineering applications. In response to the difficulty of isolating slow drift faults in dual redundancy INS through mutual comparison,a dual redundancy inertial navigation fault diagnosis method based on UAV motion model is proposed.The UAV is locally modeled based on telemetry data and aerodynamic parameters,and a difference accumulation comparison algorithm is introduced.A fault diagnosis system for dual redundancy inertial navigation is built,which is embedded in ground capture software to achieve diagnosis and isolation of slow drift faults in inertial navigation.In addition,combined with the results of an expert diagnosis system based on fault trees,the fault status of inertial navigation can be monitored in real-time in ground assist software;When a malfunction occurs in a certain inertial navigation system,the ground assist software can give a real-time alarm and assist the operator in quickly and accurately locating and determining the type of fault through significant prompts,and making the final decision.The feasibility and effectiveness of the proposed method were verified through fault data replay experiments