基于UKF的水下航行器IMU故障检测与诊断方法研究

Research on IMU Fault Detection and Diagnosis Method of Autonomous Underwater Vehicle Based on UKF

摘要:

惯性测量单元（IMU）作为水下航行器导航系统关键传感器，其可靠性直接影响航行器的导航性能。为了提高IMU的容错能力，本文提出一种基于无迹卡尔曼滤波（UKF）算法的IMU故障诊断技术。首先根据水下航行器的动力学方程和导航系统特点，建立描述IMU故障与导航状态量关系的解析模型；接着基于UKF非线性滤波的特点，进行导航滤波解算，基于此，提出了解耦矩阵法以实现IMU的故障检测；并且根据无迹卡尔曼滤波器新息正交原理，提出了实时估计IMU故障的方法，从而完成水下航行器IMU故障的在线检测与诊断。最后，通过实际航行数据验证了所提出算法的有效性。

Abstract:

The inertial measurement unit (IMU) is the key sensor of autonomous underwater vehicle navigation system, and its reliability directly affects the navigation performance of the vehicle. In order to improve the fault tolerance of IMU, this paper proposes an IMU fault diagnosis technology based on unscented Kalman filter (UKF) algorithm. Firstly, according to the dynamic equation of autonomous underwater vehicle and the characteristics of the navigation system, an analytical model describing the relationship between the IMU fault and the navigation state is established. Then, based on the characteristics of the UKF nonlinear filtering, the navigation filtering solution is performed. Based on this, the decoupling method is proposed to realize the fault detection of the IMU. And according to the orthogonal principle of UKF, the method of estimating the IMU fault in real time is proposed to complete the online detection and diagnosis of the IMU fault of the underwater vehicle. Finally, the effectiveness of the proposed algorithm is verified by the actual navigation data.