Detection of the landing site and development of algo-rithms for automatic landing of an unmanned aerial vehi-cle

Annotation: In the article, for a dual-circuit navigation and control system (NKS) of an unmanned aerial vehicle (UAV), the following tasks are solved: detection (recognition) of a landing pad (LP) for a UAV using a convolutional neural network; determining the coordinates of the LP in the frame of the UAV camera; tracking the LP in the frame through the channels of roll, yaw and pitch of the UAV triaxial gimbal; simulation of the UAV flight in space along a given trajectory and automatic landing of the UAV on the detected LP; synthesis of sliding control by the positional and angular coordinates of the UAV. MATLAB SIMULINK software was used to implement the developed algorithms for solving these problems. The created algorithms provide automatic landing of the UAV on the detected LP.

Keywords: navigation and control, neural network, landing pad, unmanned aerial vehicle, detection, automatic landing, simulation