Vision-based control and guidance structure for automatic landing simulations

In order to increase aircrafts' flight safety during near-ground operations, vision-based guidance, navigation and control solutions are investigated. Considering the hazard of testing such techniques during the approach and landing phases, simulations have to be performed prior to flight experimentations. This paper presents vision-based control and guidance structure for automatic landing simulations, which is used to simulate the approaches of automatic landings. Different guidance strategies are utilized in the simulations, based on GPS data, measurements from ILS sensors, or on camera data. Several visual features are tested and compared. Their high gains close to the runway are mitigated with a barometric altitude term. All the different strategies are evaluated in a Monte-Carlo simulation and compared with high accuracy requirements.