Unmanned Miniature Air Vehicle Control System Design

In recent years, unmanned air vehicles(UAV) have become integrated into various sectors of society. These vehicles are deployed in military operations and employed in the civilian sector for commercial and research purposes. Furthermore, distinctions may be made between different classes of UAV, in particular we focus our research here on a subclass of UAVs known as miniature unmanned air vehicles(MAV).Here we investigate the design of MAV control systems for sustained flight in the presence of wind disturbances; we are primarily focused on the low-level autopilot used to maintain parameters such as altitude, airspeed, course angle and a state estimate system of the aircraft. We consider only fixed-wing flying machines, using the MAV "Aerosonde" as the basis for our model. Our model accounts for: the choice of coordinate systems; the description of the kinematics and dynamics of the aircraft; the aerodynamic forces and moments acting on the vehicle; as well as modeled wind disturbances. Based on these particular equations, simulated using Matlab and Simulink program environments, we developed a flight model that allows us to evaluate the flight trajectory and the basic parameters of an MAV in real-time. To simplify the calculation of the lowlevel autopilot, on the basis of the balanced flight modes calculation variant we compute the transfer function of the decoupled and linearized dynamics and investigate the stability of obtained transfer functions for longitudinal and lateral movements.Regarding the low-level autopilot, a successive loop closure method was used, with saturation and the degree of intercoupling between inner control loops and outer control loops taken into consideration. In order to successfully employ this method we provide the choice basis for the structure of each control loop. Additionally, a method based on the root quality parameters is proposed here to determine the controllers gain, the application of which, guarantees stability and the necessary dynamics of the system.In order to develop the MAV’s state estimation system we established mathematical models of the pressure sensors, angular rate sensors, accelerometers, and GPS. Continuous-discrete extended Kalman filter(EKF) was designed and configured for pitch and roll angles and GPS sensor data(the sensor accounted for north and east coordinates, yaw and course angles, north and east components of the wind, and groundspeed). To determine airspeed, altitude and angular velocities a low-pass filter were calculated and configured. After explanation of the estimation algorithms, the simulation of the MAV with estimation system will be presented.Finally, we will present the simulation and the results of the estimation and control algorithms if the low-level autopilot on an imitation model aircraft. Lowlevel autopilot based successive loop closure principle and estimation method based on continues-discrete EKF has found to be reliable alternative to other, not only in flight mode without disturbance, but also in flight mode with different wind disturbance.