Spatially distributed and simultaneous wind measurements with a fleet of small quadrotor UAS

Abstract

The understanding of micro-scale flow in the atmospheric boundary layer is one major challenge in wind energy research. Besides the broad possibilities of numerical simulations, experimental data are necessary for tests of the flow conditions within a wind farm under real conditions. In wind energy and atmospheric science, a variety of measurement devices exist for measuring the wind speed. We propose a measurement system that enables completely flexible simultaneous wind measurements using a fleet of multirotor unmanned aircraft systems (UAS). This approach is validated through a two-week measurement campaign at the boundary layer field site Falkenberg of the German National Meteorological Service (DWD). The wind speed is calculated from UAS motions in hover state without additional wind sensors. The measurements are calibrated and validated against sonic anemometers mounted at a 99 m mast. The capability of highly accurate spatial distributed wind measurement with an improved wind algorithm is proven by a root mean square error (RMSE) of 0.25 ms−1 for the horizontal wind speed and < 5° for the wind direction. Further, turbulence measurements are presented showing valid results up to a frequency of 2 Hz in high turbulence conditions. Additionally, spatially horizontal distributed measurements with multiple UAS are examined in a case study of a gust front event.

Analyses of Spatial Correlation and Coherence in ABL flow with a Fleet of UAS

“…The accuracy of the wind direction measurements is determined to 4.5 • . In Wetz and Wildmann (2022) we also found that the temporal resolution of wind measurement is limited to 2 Hz due to sensor noise which dominates in small scales. This is crucial at low turbulence conditions with low signal-to-noise ratio.…”

“…We have previously shown that the calibration results for the mean wind vector and the velocity variance of the UAS measurements are in good agreement with the reference measurements using the calibration pattern close to the 99-m mast (Wetz and Wildmann 2022). The following analysis are based on the 'horizontal pattern' as illustrated in Assuming that eddies are transported with the mean advection velocity ū, the wavelengths are defined as…”

“…Since a high accuracy was reached for the mean wind speed and the variance measurements(Wetz and Wildmann 2022), TI estimates are also quite accurate with an RMSE of 3 % and an R-value of 0.96. The comparatively large error bars for high turbulent intensities are due to low wind speeds and the corresponding larger relative wind speed uncertainty, which are present for cases of TI > 0.25.…”

“…In particular, we operated our fleet at the Boundary Layer Field Site (Grenzschichtmessfeld, GM) in Falkenberg which includes various atmospheric measurement devices at the site. The most relevant for our purpose is the 99 m mast which is indicates multiple UAS in a line in front of the mast at the corresponding sonic measurement heights in 90 m and 50 m. This pattern was mainly used for calibration and validation purposes as described in Wetz and Wildmann (2022).…”

“…In Wetz and Wildmann (2022) the calculation of the horizontal wind vector used in this study is described in detail. The algorithm is based on the principle of aerodynamic drag and relates the horizontal forces to wind speeds.…”