Enhanced fault detection and isolation in modern flight actuators

Abstract

Due to their central location in the control system, actuation systems of primary control surfaces in modern, augmented aircraft must show an increased reliability. A traditional approach is based on hardware redundancy. In this way, modern actuation systems of one single control surface consist of up to two actuators and three sensors. These different dynamic subsystems are all prone to faults themselves and can be monitored. This paper presents the setup of a fault detection and diagnosis (FDD) system to systematically detect and isolate faults in the subcomponents of an actuation system. Based on the achievable fault signature matrix of the system, a residual filter is designed using nullspace theory. The residuals of the proposed filter form the basis of the decision making process to detect an isolate the faults. The developed FDD system is implemented into a nonlinear aircraft model, allowing a profound validation of the FDD system's detection and isolation performance for different actuator and sensor fault scenarios.