Autonomous, Landmark-Based Optical Navigation for Lunar Missions

In this project a simulation framework based on an existing landmark-based optical navigation solution is proposed. A self-contained library is built based on an existing algorithm for the measurement of absolute and relative landmarks and on the Extended Kalman Filter for the spacecraft state estimation. In this work, the measurements fusion strategy of the filter is improved. When a known landmark is detected and matched in a catalog, the measurement of its position is fused with the measurement of the apparent change of rate of an unknown landmark. Until the next absolute landmark is detected and matched, the navigation filter relies only on measurements of relative landmarks to limit the growth of state covariance of the estimation filter due to external perturbations. The framework is evaluated against the orbit determination requirements of the NASA Lunar Reconnaissance Orbiter mission. The public availability of high-precision spacecraft ephemeris for the low-altitude orbit and the optical parameters of the wide angle camera that is used to map the lunar surface, makes this mission a perfect candidate to successfully evaluate this autonomous navigation solution.