Wind Estimation Using Quadcopter Motion: A Machine Learning Approach
In one sentence. An LSTM estimates atmospheric wind from quadcopter motion without a wind sensor, with lower errors than the wind-triangle method in simulated turbulence.
Abstract
In this article, we study the well known problem of wind estimation in atmospheric turbulence using small unmanned aerial systems (sUAS). We present a machine learning approach to wind velocity estimation based on quadcopter state measurements without a wind sensor. We accomplish this by training a long short-term memory (LSTM) neural network (NN) on roll and pitch angles and quadcopter position inputs with forcing wind velocities as the targets. The datasets are generated using a simulated quadcopter in turbulent wind fields. The trained neural network is deployed to estimate the turbulent winds as generated by the Dryden gust model as well as a realistic large eddy simulation (LES) of a near-neutral atmospheric boundary layer (ABL) over flat terrain. The resulting NN predictions are compared to a wind triangle approach that uses tilt angle as an approximation of airspeed. Results from this study indicate that the LSTM-NN based approach predicts lower errors in both the mean and variance of the local wind field as compared to the wind triangle approach. The work reported in this article demonstrates the potential of machine learning for sensor-less wind estimation and has strong implications to large-scale low-altitude atmospheric sensing using sUAS for environmental and autonomous navigation applications.
Labels
- Topic
- Observations & remote sensing
- Type
- New model or method
- Method
- RNN / LSTM / ConvLSTM
- Domain
- Local / station
- Data
- Other
- Status
- Research
- Identifiers
- arXiv:1907.05720 DOI:10.1016/j.ast.2020.105699 OpenAlex:W3000351692
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.