Terrain relative localization for lunar entry, descent, and landing

Abstract

As exploration of the solar system continues, the need for the capability to land a spacecraft very accurately on a planetary body has become apparent. Due to limitations on the achievable accuracy of inertial navigation systems over long distances, relative navigation techniques are better suited to fill this need. Chief among these techniques is terrain relative navigation. Terrain relative navigation uses the surface features of a planetary body and an onboard map of the planetary surface to determine the spacecraft's position and update the inertial navigation system. One of the tasks of terrain relative navigation is terrain relative localization, which entails matching a sensor image of the planetary surface to the stored map of the surface. This correlation allows a position match to be determined, which is used to update the spacecraft's inertial navigation system. This thesis focuses upon two terrain relative localization techniques and their applicability to lunar entry, descent, and landing. These localization techniques are terrain contour matching (TERCOM) and a crater matching routine. Both methods are tested using simulation environments that mimic expected conditions in lunar navigation. The ability of each algorithm to generate a position match with no noise is evaluated, as well as the performance of each algorithm under various sensor noise conditions. The terrain contour matching algorithm generates a high level of error in the position match and is found to be unsuitable for lunar terrain relative navigation. The crater matching routine performs quite well, with low processing speeds, moderate memory requirements, and a high level of position match fidelity under moderate noise conditions. The crater matching routine is recommended for continued work and potential application to lunar navigation during entry, descent, and landing.