Autonomous Landing of Fixed-Wing UAV in a Stationary Net - Path and Navigation System
MetadataVis full innførsel
This thesis present a path and navigation system, used in a autonomousnet landing system for a fixed wing Unmanned Aerial Vehicle (UAV). Alanding path of a UAV can be constructed as a straight line path, but inorder for a UAV to follow the landing path it must be in a position fromwhich it has a feasible path to the start position of the landing path. Thismotivates the development of a approach path logic towards the landingpath entry position from any initial start position in the air space. In addition to a path generation system the UAV require a robusthigh accurate navigation system. This is accomplished by applying RealTime Kinematic GNSS (RTK-GNSS), which can provide centimeter levelposition accuracy. A shortcoming of the RTK-GNSS system is that it mayloose its lock on satellites leading to loss of functionality. In this workthis is compensated for by introducing a secondary Global NavigationSatellite System (GNSS) system. To handle a RTK-GNSS drop out arobust RTK-GNSS system is proposed, where previous valid RTK-GNSSposition solutions are fused together with the secondary GNSS system,to be used as a compensator for the external navigation system. Thecompensator is designed to enable the external navigation system toachieve the same position accuracy level as the RTK-GNSS system for ashort duration, until the RTK-GNSS is either reconnected or completelydisconnected. With the compensator the UAV navigation system becomesrobust against short drop out of the RTK-GNSS, and the availability ofthe RTK-GNSS is prolonged. Experimental testing, in addition to Software In the Loop (SIL)verification, of the ability of UAV to land in a stationary arbitrary placednet has been performed. For improved navigation and performance, anmobile sensor unit have been utilized to provide the required positiondata of the net. This sensor unit was used during testing, where it showedto provide the intended contribution.