Helicopter main rotor dynamics by neuro-fuzzy classification with bred vector

Helicopters are aircraft with rotational wings with many sub-systems presenting nonlinear dynamics. One of these sub-systems is the main rotor, which is responsible to provide the sustentation of the aircraft. Nonlinear equations of motion are applied to simulate the dynamics of the main rotor. The helicopter dynamics can presents a very dangerous situation, such as the ground resonance caused by the interaction between the ground and the rotorcraft. The rotor dynamical analysis is performed by breeding technique, where the model is executed under a perturbed initial condition. The difference between the reference and the perturbed dynamics is called bred vector. The time series of bred vector magnitude is employed for describing classes of dynamics predictability, in other words, the classes represent the degree of confidence for the simulated dynamics. Two neuro-fuzzy classifiers are used in our analysis: ANFIS and GUAJE software packages.