A Robust In-Motion Alignment Method With Inertial Sensors and Doppler Velocity Log
Xiang Xu, Jing Gui, Yifan Sun, Yiqing Yao, Tao Zhang
Abstract
The initial alignment process is of vital importance to the strapdown inertial navigation system (SINS), and SINS is consisting of the inertial sensors. In this article, the aiding velocity of Doppler velocity log (DVL) is used to implement the in-motion alignment process. However, the measurements of DVL are easy to corrupt by the outliers, and the outliers will degrade the performance of the initial alignment process. Thus, a robust in-motion alignment method for the SINS/DVL integrated navigation system is proposed in this article. First, the vector construction with the measurements of SINS and DVL is introduced. Then, the discrete form of the vector by considering different sampling rates between SINS and DVL is derived in detail. By analyzing the relationship between observed and reference vectors, the magnitude matching method is devised for outliers detecting. Meanwhile, Huber’s robust theory is used to calculate the weighted value with the magnitude matching method. Using the weighted value, the filter-QUEST attitude determined method is modified to implement a robust attitude determination. Finally, the simulation and field tests are designed to validate the performance of the proposed method. The alignment results demonstrate that the proposed method has similar accuracy with the current popular method in which the measurements of DVL are not corrupted by the outliers.