Wind Field Estimation Research Based On Quadrotor UAVs

Nowadays,people pay more attention on the unmanned aerial vehicles(UAVs)for its application in atmospheric measurement and collection.Compared to the fixed wing UAVs,unmanned rotorcrafts take the advantages of flying at low altitude,hovering at a certain position and working in dangerous environments,which are regarded as ideal platforms for wind field estimation.Therefore,this paper proposes a quadrotor-based approach for wind field measurement and retrieval without additional meteorological sensors.Meanwhile,this method provides an online solution with high spatial and temporal resolution for the wind estimation near ground.The main contributions are shown as follow.(1)The dynamics of a quadrotor UAV is investigated with the influence of wind.Firstly,the quadrotor's dynamic model is obtained according to the Newton-Euler equations.Then,in order to describe the model accurately,a research on the rotor's aerodynamics is conducted based on the momentum theory and the blade element theory.In the study,the rotate speed,wind speed and airflow angle are taken into consideration,which have an influence on the rotor's force and torque.Hence,the accurate model of quadrotor is achieved for hovering control law design and wind estimation research.(2)A hovering control law is designed for the quadrotor to reject wind disturbance.Considering the requirement about fast performance,none static error and noise rejection,the quadrotor's attitude tracking controller is designed based on the adaptive sliding mode methods.Furthermore,due to the complexity of the cascade control system,a PID control law is utilized in the position loop.Thus,the quadrotor UAV are able to hover in a windy environment,which provides a platform for wind field estimation.(3)Two lateral wind estimation methods are presented on the quadrotor-based platform.One inverses the model of the quadrotor to get the expressions of the lateral wind,relying on the quadrotor's dynamics.Another is on the basic of the relationship between the bank angle and the wind vector.In the simulation,the wind estimating results are analyzed on the mean wind,turbulence and gust,which are classified as the classical wind environment at low altitude near ground.(4)Two spacial wind retrieval algorithms are proposed with multiple samples on the quadrotor-based platform.These approaches reconstruct the wind field,depending on the single factor regression model and Kalman filtering regression model,respectively.In the simulation,a prevailing wind estimate algorithm is firstly employed based on the Kalman filtering method.Then,the proposed approaches are used to retrieve the shear wind,which verifies the effectiveness of the algorithms.(5)The experimental scheme is designed about wind field standardization.Through the indoor and outdoor experiments,the relationship between the bank angle of the quadrotor and the wind vector will be achieved,which can be used for the practical application.The validity of the testing program is verified by the flight data.