M.F. Al-Malki and D.-W. Gu (UK)
Fault detection and Isolation (FDI), Helicopter DynamicsModelling, Helicopter sensors, Artificial neural networks(ANNs).SymbolsSignals:FDA final drives to actuators in inches ofSafety Pilo
In this paper, we report the achieved results on sensors' faults detection and isolation for a Bell-205 Helicopter which is not equipped with any physical sensor redun dancy. The helicopter dynamics is split into lateral and longitudinal channels. Artificial neural network (ANN) models have been trained using real flight data, which contributes to the robustness of models. The simulation results show that the neural network models are capable to capture the characteristics of the lateral and longitudinal dynamics with high degree of accuracy. Neural network models trained in various flight conditions results in successful fault detection and isolation.
Important Links:
Go Back
