Discrete-time observers for inertial navigation systems

Abstract

In this thesis, we derive an exact deterministic nonlinear observer to compute the continuous-time states of inertial navigation system based on partial discrete measurements, the so-called strapdown problem. Nonlinear contraction theory is used as the main analysis tool. The hierarchical structure of the system physics is sytematically exploited and the use of nonlinear measurements, such as distances to time-varying reference points, is discussed. Effects of bounded errors on model and measurements are quantified, and can be used for active measurement selection. Work on vehicle state computation is carried out by using a similar observer design method. Finally, the approach is used to compute the head orientation of a simulated planar hopping robot, where the information provided by the observer is used for head stabilization and obstacle jump.