The use of Autonomous Underwater Vehicles (AUV) as robots for exploration and oceanology science has been a field of interest of several universities and research centres around the world in the last decade. In particular, autonomous navigation, localization and mapping have become topics of increasing interest both for researchers in underwater robotics and marine scientists. Nevertheless the performance of AUV navigation, is still limited by the precision and accuracy of the navigation solution deployed on it. Inertial Navigation (INS) has commonly been used as the main means of localization for autonomous underwater vehicles. The main disadvantage in the use of INS (particularly when it is implemented with low-cost sensors), is the growing error in the estimated positions due to the dead-reckoning nature of the method. Traditionally\ external sensors such as: GPS, transponders from a time of flight acoustic navigation method, Doppler velocity log (DVL), sonar or a camera have been used to aid the solution of the inertial navigator, thus constraining the errors in the estimated positions. These external sensors have several practical disadvantages, basically related to the reliance on external information, such as reception of satellite transmissions or reliable observable ocean floor. One source of information that can assist in the localization of the vehicle, without the need for extra additional external sensing, is the use of its Dynamic Model. The model of the vehicle is capable of representing the attitude of the system according to the control inputs and the external forces acting on it.
|Date of Award||2015|
|Supervisor||Hichem Sahli (Promotor) & Hernandez Luis (Promotor)|
- System identification
- kaalman filter