Research on fault detection technology of autonomous underwater vehicle navigation system

In order to improve the reliability and high precision of AUV integrated navigation system in the complex Marine environment, the strap-down inertial navigation system is adopted as the reference system of AUV integrated navigation system, and the Doppler velocity sonar and global positioning system are used as the auxiliary navigation system to form the integrated navigation system In order to solve the problem that the classical residual chi-square fault detection method is difficult to detect the gradual fault, a residual fault detection method based on vector normalization is proposed The simulation results show that the residual fault detection based on vector normalization is more timely than the residual Chi-square fault detection method in detecting gradual faults, which is conducive to the timely isolation and processing of the autonomous underwater vehicle navigation system and improves the reliability of the navigation system.