Wind field estimation through autonomous quadcopter avionics

Abstract

Using integrated sensors carried onboard a quadrotor unmanned aerial vehicle (UAV) can be used to wind field estimation through intelligent dynamic analysis, UAV control, and sensor management. The data from UAV on-board sensors such as GPS and inertial measurement units are utilized such that no dedicated sensors (i.e., pitot tube) for wind characterization are necessary. Using the estimated ground weather conditions, the UAV performance provides a means for rapid wind field estimation. The motivation is to develop an agile and low-cost atmospheric measurements system for energy harvest and realtime mission support. The advantage of UAV versus weather balloons is the agility of the UAV to operate in constrained environments and complex terrain. The wind profile is calculated by applying algorithms that relates the attitude of the aircraft to the local wind speed and direction, sparing the payloads of external devices like multi-hole tubes. Several existing wind field estimation algorithms are evaluated and compared with the proposed Kalman Filter dynamic behavior fusion using data obtained from the wind sensors as well. Error analysis and the reasons of errors are discussed in the estimation process. The proposed UAV-based on-board avionics system can be used to improve positioning accuracy and flight stability in the spatially varying, turbulent wind circumstances.