Research On Control System Design Of Auto Landing For Unmanned Aerial Vehicle

The flight course of UAV can be divided into three segments: ground taxiing and taking off,in-flight and landing. It is necessary to take full account of differences in its three phases whenestablishing the mathematical model of the wheel UAV. To facilitate the controller design, the wholeamount models are trimly linearized. Combined with the existing aerodynamic data, structuralparameters, etc, this thesis analyzes the characteristics of the sample UAV, such as the lift-drag ratiocharacteristic, static stability and the vertical and horizontal stability.To ensure the landing safety of the sample UAV, the movement rate should be kept within acertain range and the UAV should keep its head lifting when landing. Therefore, a glide trajectorybased on the distance is designed for the sample UAV. This glide trajectory includes approach flightpath line segment, trajectory capture segment, plummeting segment, pull up segment, and the groundtaxiing segment. The exact control parameters of each glide trajectory segments is decided byindicators of the sample UAV, including the minimum&maximum speed, maximum thrust, altituderange, angle of attack, the scope of work and etc.The control law and control parameter of the sample UAV in section of landing are importantfactors which affect the landing safety. Based on the analysis of the vertical and horizontalcharacteristics, this thesis designs the control law and select the control parameters of the verticalcontrol loop and the cross-lateral control loop respectively. The vertical control loop includes heightcontrol loop and pitch attitude control loop. The cross-lateral control loop includes roll attitudecontrol loop, heading control loop and track control loop. The simulation results show that the designof control laws can fully meet the requirements of the sample UAV landing.Accurate navigation results are essential to the accurate landing of UAV. This thesiscomprehensively considers the factors that the flight environment to the navigation sensors, andproposes an integrated GPS/SINS/Vision navigation system aided by barometric altimeter and radioaltimeter in the landing segment. Firstly the working principles and error models of the GPS, SINS,barometric altimeter and radio altimeter are analyzed, then the working principle of a ground visualnavigation system is emphasized. The federated kalman filter algorithm is used to handle theinformation from each sensor. An adaptive information sharing coefficient based on the eigenvalue ofeach sub-filter’s covariance matrix is designed to solve the shortage of the traditional fixedinformation sharing coefficient. Finally, the experimental simulation verifies that the integrated navigation system proposed in the thesis can provide accurate navigation results, thus meet theaccuracy requirements which the landing of UAV make of the navigation system.