Floor optical flow based navigation controller for multirotor aerial vehicles

Abstract

The presented work is focused on GPS-denied autonomous navigation for multirotor platforms. The problem of safe navigation based on noisy odometry measurements is discussed, and experimentally tested on the case of on-board optical flow measurements. A navigation controller is proposed which allows, through a group of configurable parameters, to ensure that the vehicle will fly on a speed specified flight envelope where the quality of the optical flow measurements is guaranteed. In order to attain safe navigation, the multirotor is modeled as a flying vehicle with specific kinematic constraints. The designed unperfect odometry based controller architecture has been experimentally tested on various multirotor vehicles, where the vehicles featured similar sensoring capabilities and the tolerance of our approach have been demonstrated. This work was implemented to compete in the International Micro Air Vehicle Conference and Flight Competition IMAV 2012, gaining two awards: the Special Award on “Best Automatic Performance - IMAV 2012” and the second overall prize in the category of “Indoor Flight Dynamics - Rotary Wing MAV”. Most of the related code is available as two open-source projects hosted on GitHub.