Terrain Aided Recovery of a Multirotor Using Laser Range-Finders in a Satellite-Navigation Deprived Environment

Unmanned aerial vehicles rely predominantly on external signal reception for navigation. This includes both satellite navigation (GNSS) and guidance through remote control channels. In case of a signal loss, a recovery logic is needed to navigate the vehicle back to its base until normal operation is regained. Such systems commonly require pre-installed local maps of the flight area or rich-information sensors like LIDARs and cameras.

Our work examines a recovery logic for a multirotor based on several Body-Fixed Laser Range-Finders (BF-LRFs) and an inertial measurement unit (INS) without a pre-installed map, hence providing a simple, low-cost recovery solution. BF-LRFs were shown to perform well for Terrain Following for single UAVs and multirotor slung load applications. To facilitate the recovery, a terrain map along the flight trajectory can be constructed during the normal GNSS operation when an accurate navigation solution is available using the BF-LRFs as terrain sampling sensors. In case of GNSS signal loss, the accumulated map, together with an INS and the BF-LRFs measurements, are used for a terrain-aided navigation type solution to navigate the vehicle back to its base. Along with the recovery system, a BF-LRF based obstacle avoidance solution is examined both separately from the recovery system and combined. The recovery logic was tested in a realistic six-degrees-of-freedom simulation using a real terrain map. The simulations evaluated the recovery and the obstacle avoidance systems in several different LRF configurations and flight paths, showing promising results.