Development of position estimation algorithm for unmanned aerial vehicle at short term GPS outage

Last years, unmanned aerial vehicles (UAVs) are used in many areas such as mapping, search-and-rescue and visualization in civil and millitary applications. The number of companies that produce UAV increases with the expansion of their usage areas day by day. Parallel to this, UAV has reached large number of users. Also, last years the number of researchers work on UAV is increased. UAVs became very important for defense industry applications because they are autonomous and controlled remotely. Thus, the demand of UAV in terms of defense and security purposes in our contry and in the world is quite high. As increasing usage of UAVs in defense industry, the importance of complete their mission is vital. Also, UAVs should be durable to harsh environmental conditions. UAVs have many features such as hovering, go to targets, travelling between targets or go to home. With increasing demand and field of use of UAVs, new main desired fatures are increasing. Thus, these vehicles need some variaty of sensors, sensor filter, sensor fusion and fault tolerant algorithms. Additionally, sometimes sensors may be temporary or permanent outages due to hardware problems and environmental effects. In this outage scenarios, UAV must not lose its control to avoid a fall. UAVs which can be producted different shape and size can be controlled remotely or autonomous with the help of ground control station. UAVs need well-designed controllers to operate their task safety and avoid destructive effects. Sensor noieses may cause unstability. Because of that, real time controller and filter structure are used for system security. Moreover, sensor error and fusion algortihms have to be developed. The vehicle's rotational angles, angular velocities and accelerations are obtained with the help of the inertial measurement unit (IMU) sensor on the vehicles. Vehicle position is obtained with global positioning system (GPS), and vehicle speed can be measured with position changes. In this thesis, posisiton estimation algorithms were developed and implemented to handle short term GPS outages. GPS signals can be interrupted because of surronding hills, buildings or tree-lined environments. Also, as these vehicles are used for millitary applications, man-made jammers can be used to reduce GPS signals. In this situations, vehicle speed and position can be measured by vehicle accerations and oriantation angles which are obtained by others sensors. Sensors errors and noise should be filtered with low pass, high pass or kalman filters to estimate speed and position at short term GPS free enviroments. Also, wind estimation was made within this thesis. The developed algorithms which are described in this thesis have been verified with flight tests. According to the test results, one of the algorithm was chosen and was tested with GPS jammer. Then, GPS jammer was used during second tests to interrupt GPS signals. Test results and algorithm performance were evaluated within this study.