This paper contains a concise comparison of a number of nonlinear attitude filtering methods that have attracted attention in the robotics and aviation literature. With the help of previously published surveys and comparison studies, the vast literature on the subject is narrowed down to a small pool of competitive attitude filters. Amongst these filters is a second-order optimal minimum-energy filter recently proposed by the authors. Easily comparable discretized unit quaternion implementations of the selected filters are provided. We conduct a simulation study and compare the transient behaviour and asymptotic convergence of these filters in two scenarios with different initialization and measurement errors inspired by applications in unmanned aerial robotics and space flight. The second-order optimal minimum-energy filter is shown to have the best performance of all filters, including the industry standard multiplicative extended Kalman filter (MEKF).
On Fast Attitude Filtering Using Matrix Fisher Distributions with Stability Guarantee as an Observer
School of Astronautics, Shenyuan Honors College, Beihang University（北京航空航天大学宇航学院、深空探测学院）；School of Astronautics and the MOE Key Laboratory of Spacecraft Design, Optimization and Dynamic Simulation, Beihang University（北京航空航天大学宇航学院、航天器设计与优化动态仿真教育部重点实验室）；School of Astronautics, Beihang University（北京航空航天大学宇航学院）
Complementary Filtering on SO(3) for Attitude Estimation with Scalar Measurements
I3S, CNRS, Université Côte d’Azur（法国国家科学研究中心蔚蓝海岸大学I3S实验室）；Institut Universitaire de France（法国大学研究院）；Department of Computer Science and Engineering, Université du Québec en Outaouais (UQO)（魁北克大学乌塔韦校区计算机科学与工程系）
Multi-Robot Relative Pose Estimation in SE(2) with Observability Analysis: A Comparison of Extended Kalman Filtering and Robust Pose Graph Optimization
