Rigid body motions estimation using inertial sensors: Bio-logging application

Bio-logging is a new interdisciplinary research area at the intersection of animal behavior and bioengineering. It can involve several applications such as determination of specific parameters measurements (attitude, acceleration, and position) via a new generation of adapted-animal mechatronic systems (a.d.a.m.s.). The aim of this paper concerns the animal motion estimation problem using low-cost Micro-Electro-Mechanical-System (M.E.M.S.) sensors fusion. Generally, accurate attitude measurements are obtained using low-pass sensors such as accelerometers and magnetometers, but at quasi-static motion. In this work a state estimation algorithm is proposed, it is based on a combination of an observer and an iterated Least Squares Algorithm (I.L.S.A.). This approach fuses data from 3-axis accelerometer, 3-axis magnetometer, and 3-axis gyroscope to provide the best attitude estimation when the rigid body is both in static and dynamic motion. A correction stage is introduced to take into account the linear acceleration effect undergone by the rigid body. Finally, some simulation and preliminary experimental results are provided to show the efficiency of the proposed algorithm.