Several and distinct Unmanned Aircraft Vehicle (UAV) applications are emerging, demanding steps to be taken in order to allow those platforms to operate in an un-segregated airspace. The key risk component, hindering the widespread integration of UAV in an un-segregated airspace, is the autonomous component: the need for a high level of autonomy in the UAV that guarantees a safe and secure integration in an un-segregated airspace. At this point, the UAV accurate state estimation plays a fundamental role for autonomous UAV, being one of the main responsibilities of the onboard autopilot. Given the 21st century global economic paradigm, academic projects based on inexpensive UAV platforms but on expensive commercial autopilots start to become a non-economic solution. Consequently, there is a pressing need to overcome this problem through, on one hand, the development of navigation systems using the high availability of low cost, low power consumption, and small size navigation sensors offered in the market, and, on the other hand, using Global Navigation Satellite Systems Software Receivers (GNSS SR). Since the performance that is required for several applications in order to allow UAV to fly in an un-segregated airspace is not yet defined, for most UAV academic applications, the navigation system accuracy required should be at least the same as the one provided by the available commercial autopilots. This research focuses on the investigation of the performance of an integrated navigation system composed by a low performance inertial measurement unit (IMU) and a GNSS SR. A strapdown mechanization algorithm, to transform raw inertial data into navigation solution, was developed, implemented and evaluated. To fuse the data provided by the strapdown algorithm with the one provided by the GNSS SR, an Extended Kalman Filter (EKF) was implemented in loose coupled closed-loop architecture, and then evaluated. Moreover, in order to improve the performance of the IMU raw data, the Allan variance and denoise techniques were considered for both studying the IMU error model and improving inertial sensors raw measurements. In order to carry out the study, a starting question was made and then, based on it, eight questions were derived. These eight secondary questions led to five hypotheses, which have been successfully tested along the thesis. This research provides a deliverable to the Project of Research and Technologies on Unmanned Air Vehicles (PITVANT) Group, consisting of a well-documented UAV Development and Validation of an IMU/GPS/Galileo Integration Navigation System for UAV II navigation algorithm, an implemented and evaluated navigation algorithm in the MatLab environment, and Allan variance and denoising algorithms to improve inertial raw data, enabling its full implementation in the existent Portuguese Air Force Academy (PAFA) UAV. The derivable provided by this thesis is the answer to the main research question, in such a way that it implements a step by step procedure on how the Strapdown IMU (SIMU)/GNSS SR should be developed and implemented in order to replace the commercial autopilot. The developed integrated SIMU/GNSS SR solution evaluated, in post-processing mode, through van-test scenario, using real data signals, at the Galileo Test and Development Environment (GATE) test area in Berchtesgaden, Germany, when confronted with the solution provided by the commercial autopilot, proved to be of better quality. Although no centimetre-level of accuracy was obtained for the position and velocity, the results confirm that the integration strategy outperforms the Piccolo system performance, being this the ultimate goal of this research work.    «

Several and distinct Unmanned Aircraft Vehicle (UAV) applications are emerging, demanding steps to be taken in order to allow those platforms to operate in an un-segregated airspace. The key risk component, hindering the widespread integration of UAV in an un-segregated airspace, is the autonomous component: the need for a high level of autonomy in the UAV that guarantees a safe and secure integration in an un-segregated airspace. At this point, the UAV accurate state estimation plays a fund...    »