Unmanned Aerial Vehicles (UAV) are commonly used for robotics research and industrial purposes. Most of the autonomous applications use visual sensors and inertial measurement units for localization. Design constraints of such systems are defined considering smooth operation requirements such as indoor environments without external forces where input tracking signal is constant during an operation. In this research paper, we simultaneously investigate and compare stability, power consumption and landmark tracking quality of a visual sensor mounted gimbal specifically for rapid UAV motion requirements where input signal continuously varies such as at obstacle rich environments. We not only attempt to find efficient control parameters but also compare these settings with power consumption and landmark tracking quality metric which are vital for mobile robots and localization algorithms. Efficiency of the system response is analyzed with rise and settling time as well as oscillation amplitude and frequencies. These parameters are tested and benchmarked with various voltage and current limitations. In addition to that, different response behaviors were investigated considering landmark tracking quality metrics including feature detection and image blur. We have shown that gimbal stabilization controller under continuously varying input signal requires less responsive behavior to keep landmark tracking accuracy stable. Initial simulation results, system development and experimental setup procedure are explained and behavior plots for each topic are listed and analyzed.

This content is only available via PDF.
You do not currently have access to this content.