Date of Award
Master of Science
Greg R. Luecke
A large number of attitude estimation algorithms assume sensor placement along the axis of rotation. These models assume that all measured accelerations are due to gravity, while in reality this is not always the case. By expanding the system model to accommodate centrifugal and tangential accelerations the attitude can be more accurately estimated. This thesis introduces an approach to model accelerations not caused by gravity.
This model is incorporated into a Kalman filter and then compared with the more conventional Kalman filter approach which assumes all system acceleration is due to gravity.
Gartner, Keegan, "Improved rate and angle estimation through higher accuracy planar motion models" (2011). Graduate Theses and Dissertations. 10359.