Aerodynamic Modeling using Neuro-Fuzzy for a scaled model of Cessna 182

Abstract

This work has the main objective to obtain an aerodynamic model, as fidelity as possible, for anaircraft remote controlled, a replica of the Cessna 182 in scale. To conduct the idea, was used thecomputational intelligence technique known as Neuro-Fuzzy, which is capable of imitating thebehavior of any system, if constructed with the help of a specialist. The experimental data wereacquired through an acquisition board known as Pixhawk-PX4. After data acquisition, the samewas divided into two parts, training and validation. After the optimization made by Neuro-Fuzzy,with the aid of a differential evolution for the training data, the model should be able to predictthe forces and moments for different attitudes of the air model.