Fixed Wing UAV Navigation and Control Through Integrated GNSS and Vision

Abstract

With the rapid deployment of Unmanned Airborne Vehicles (UAVs) into new applications, the pressure to extend the capabilities of current platforms is increasing. Increased capabilities, however, should preferably not come at the cost of increased aircraft size. In order to strive towards a more capable platform, the UAV must become increasingly aware of its current state (control, navigation and health) and surroundings (location of other aircraft, airspace boundaries, weather and terrain). This paper reports on the results of research into providing a new level of situational awareness to the UAV that is low in cost and complexity. In particular the paper investigates the unique benefits that can be obtained from the integration of a GNSS sensor and a forward-looking vision sensor. The motivation for this investigation is the belief that both GNSS and vision will be integral features of future UAV avionics architectures: GNSS to provide basic aircraft navigation; and vision to provide for obstacle, and aircraft collision avoidance. This paper will present results showing that when single-antenna GNSS measurements are combined with information derived from optical flow techniques, a number of unique synergies emerge. Sensor accuracies and simulated flight control results are presented based on a comprehensive Matlab® Simulink® model which creates an optical flow stream based on the simulated flight of an aircraft. The paper establishes the promise of this novel integrated GNSS/Vision Sensor Suite approach for use as a complete UAV sensor package, or as a backup sensor for an inertial navigation system.