Heading Bias Estimation Algorithm for Autonomous Underwater Vehicles Considering Angular Accuracy

This paper describes a method for estimating the heading bias to improve the performance of the navigation algorithm for AUVs (Autonomous Underwater Vehicles). Various studies have been conducted to estimate heading bias, including approaches that use position data obtained from sensors providing absolute position information, such as the GPS (Global Positioning System), along with position data estimated through navigation algorithms. However, these studies have estimated bias without considering that the reliability of each GPS position varies depending on its relative location. Therefore, this paper proposes a method for heading bias estimation that involves estimating angular accuracy based on the distance between the initial position of bias estimation and GPS positions, GPS accuracy, and distance-to-angle conversion. Based on the estimated angular accuracy, GPS positions for bias estimation are selected, and the heading bias is estimated by applying optimization techniques to these selected GPS positions and their associated navigation-estimated positions. The effectiveness of the proposed algorithm is validated through experiments using sensor datasets acquired from real sea experiments.