VALIDATION OF INERTIAL AND OPTICAL NAVIGATION TECHNIQUES FOR SPACE APPLICATIONS WITH UAVS

摘要

PERIGEO is an R&D project, funded by thernINNPRONTA 2011-2014 programme from SpanishrnCDTI, which aims to investigate the use of UAVrntechnologies and processes for the validation of spacernoriented technologies. For this purpose, among differentrnspace missions and technologies, a set of activities forrnabsolute and relative navigation are being carried out torndeal with the attitude and position estimation problemrnfrom a temporal image sequence from a camera on thernvisible spectrum and/or Light Detection and Rangingrn(LIDAR) sensor. The process is covered entirely: fromrnsensor measurements and data acquisition (images,rnLiDAR ranges and angles), data pre-processingrn(calibration and co-registration of camera and LIDARrndata), features and landmarks extraction from thernimages and image/LiDAR-based state estimation.rnIn addition to image processing area, classicalrnnavigation system based on inertial sensors is alsornincluded in the research. The reason of combining bothrnapproaches is to enable the possibility to keeprnnavigation capability in environments or missions wherernthe radio beacon or reference signal as the GNSSrnsatellite is not available (as for example an atmosphericrnflight in Titan). The rationale behind the combination ofrnthose systems is that they complement each other. ThernINS is capable of providing accurate position, velocityrnand full attitude estimations at high data rates. However,rnthey need an absolute reference observation torncompensate the time accumulative errors caused byrninertial sensor inaccuracies. On the other hand, imagingrnobservables can provide absolute and relativernpositioning and attitude estimations. However they needrnthat the sensor head is pointing toward groundrn(something that may not be possible if the carryingrnplatform is maneuvering) to provide accuraternestimations and they are not capable of provide somernhundreds of Hz that can deliver an INS. This mutualrncomplementarity has been observed in PERIGEO andrnbecause of this they are combined into one system.rnThe inertial navigation system implemented inrnPERIGEO is based on a classical loosely coupledrnINS/GNSS approach that is very similar to thernimplementation of the INS/Imaging navigation systemrnthat is mentioned above.rnThe activities envisaged in PERIGEO cover thernalgorithms development and validation and technologyrntesting on UAVs under representative conditions. Pastrnactivities have covered the design and development ofrnthe algorithms and systems. This paper presents thernmost recent activities and results on the area of imagernprocessing for robust estimation within PERIGEO,rnwhich are related with the hardware platforms definitionrn(including sensors) and its integration in UAVs. Resultsrnfor the tests performed during the flight campaigns inrnrepresentative outdoor environments will be alsornpresented (at the time of the full paper submission therntests will be performed), as well as analyzed, togetherrnwith a roadmap definition for future developments.