AUTOMATION OF SMALL UAVS USING A LOW COST MEMS SENSOR AND EMBEDDED COMPUTING PLATFORM

摘要

This paper presents a framework for the automation of a small UAV using a low cost sensor suite, MNAV, and an embedded computing platform, Stargate, which together provide a complete avionics package for aerial robotic applications. In order to provide a complete INS solution (i.e., attitude, velocity, position, and biases), an extended Kalman filter algorithm is developed and implemented in real-time. A devised control strategy utilizes multiple PED loops with a hierarchy enabling simple attitude stabilization to full waypoint navigation. The developed ground station unit, a laptop computer, communicates with the avionics package via 802.1 lb WiFi, displays the aircraft critical information, provides in-flight PID gain tunings, and uploads waypoints, through a simple GUI. The system is installed in an off-the-shelf delta-wing R/C aircraft and demonstrates its performance for aerial robotic applications.