Measuring Atmospheric Boundary Layer Profiles Using UAV Dynamic Response

摘要

In this work, an approach to predict the atmospheric wind velocity is explored using unmanned aerial vehicles (UAVs). The approach has multiple aspects that span from vehicle control to developing verification data. To provide physical atmospheric data to test our algorithms, we propose using a pitot tube with an experimental quadcopter platform. However, interactions between a pitot tube mounted on a quadcopter with respect to both the induced flow from the rotor and incidence are not fully understood. For this reason, a computational fluid dynamics (CFD) study is completed to evaluate the capability of using a pitot tube to measure local free-stream velocities. Results show that the relative angle of the pitot tube to free stream flow and low-pressure region created by the propellers affect the measurement and will likely involve a bias. Additionally, we develop a dynamics model of the quadcopter in response to wind forcing as well as its controlled response to remain near a hover equilibrium. As our goal is to measure wind speed, we utilize an inverse dynamic system to estimate the oncoming wind components of a validation oncoming wind scenario. Hence, the inverse dynamic model uses the Massey-Sain linear system inversion algorithm. In order to evaluate the linear system, synthetic atmospheric wind velocities in conjunction with a non-linear quadcopter dynamics model are used as a baseline. Results indicate promise, but a need to refine the method to obtain improved measurements.