Information Fusion Algorithm for Damping Attitude of Strapdown Inertial Attitude and Heading Reference System Based on Fault Diagnosis

摘要

Strapdown Inertial Attitude and Heading Reference System (SIAHRS) is a kind of autonomous and cost-efficient inertial navigation system (INS), which is mainly applied to output vehicle's attitude. In this paper, special attention is focused on increasing attitude accuracy with the information of inertial measuring unit (IMU) its own. In other words, we try to obtain more accurate attitude when external sensors are all unavailable in the system. It is especially important in military situation. Due to gyros' drift, there are several errors existed in INS. In traditional system of platform type, the velocity information is employed to damp the horizontal attitude circuit and suppress the vibration error. According to this idea, the paper suggests a method to calculate vehicle's attitude only by the information from accelerometers, and this attitude is named damping attitude. Kalman filter is designed to integrate damping attitude with that calculated by IMU. Because damping attitude is accurate only when the vehicle is moving without acceleration, two methods are proposed to judge vehicle's motion. One is to employ the horizontal output of accelerometers. The other is to employ residual chi-square test in Kalman filter, and proper information fusion strategy is designed as well. At last, the new algorithm above is proved by actual system's experiments.