Surface failure detection for F/A-18 aircraft using neural networks and fuzzy logic

In this paper we consider the problem of detecting control surface failures of a high performance aircraft. The detection model is developed using a linear, six degree of freedom dynamic model of an F/A-18 aircraft. The detection scheme makes use of a residual tracking error between the actual system and the model output in order to detect and identify a particular fault. Two parallel models detect the existence of a surface failure, whereas the isolation and magnitude of any one of the possible failure modes is estimated by a decision algorithm using either neural networks or fuzzy logic. Simulation results demonstrate that detection can be achieved without false alarms even in the presence of actuator/sensor dynamics and noise