In this paper, an improved residual chi-square test fault isolation approach is proposed. In order to improve reliability of strapdown inertial navigation system (SINS), redundant SINS composed of redundant inertial… Click to show full abstract
In this paper, an improved residual chi-square test fault isolation approach is proposed. In order to improve reliability of strapdown inertial navigation system (SINS), redundant SINS composed of redundant inertial sensors is applied. Among redundant SINS, four-gyro SINS has a wide range of applications as it supplies big reliability considering cost and volume. Fault isolation (FI) can isolate a fault when a fault occurs in redundant SINS. Generalized likelihood test (GLT) fault isolation based on kalman filter (KF) effectively isolate a fault because of its high sensitivity, small calculation and easy implementation. Nevertheless, GLT fault isolation based on KF cannot recognize a fault in four-gyro SINS. In this paper, first, a new observation model is introduced based on error models of a redundant SINS. In addition, a residual vector for isolation based on KF generated by gyros and star sensors is designed. Second, a separate residual chi-square test fault isolation approach for each gyro is proposed to recognize a fault. Third, an average separate residual vector and new isolation threshold are designed to reduce isolation time. At last, a star sensor is employed to obtain angular velocity, which provides angular velocity baseline information for the proposed isolation approach. Simulation shows that the proposed approach is useful to recognize a fault in four-gyro SINS.
               
Click one of the above tabs to view related content.