Wind Estimator Using Attitude Measurement From Quadrotor Flight Under Wind Disturbance

There is a limitation to flying a quadrotor in the lowest layer of the atmosphere, the troposphere level. Thus, it is difficult to evaluate the performance of the quadrotor under the presence of wind. The main objective of this project is to validate the quadrotor control performance under the proposed wind prediction model. A wind estimator model was designed using neural network models to validate the quadrotor model with a proportional integral derivative(PID) controller, flying under external disturbance. The performance of the wind estimator model was evaluated based on error measurement. Thus, the actual flight data and the estimated data were compared and evaluated to obtain the best performance for the quadrotor flight control. The simulation results of the wind estimator signified that the model has been successfully developed according to the set parameters. Thus, the outcome of this project shows that neural network fitting can be embedded inside the quadrotor and work together with the existing PID controller to control the quadrotor in a robust environment.