Robust, model-based fault detection for commercial transport air data probes.

Abstract

Air data probes provide essential sensing capabilities to aircraft. The loss or corruption of air data measurements due to sensor faults jeopardizes an aircraft and its passengers. To address such faults, sensor hardware redundancy is typically combined with a voting system to detect and discard erroneous measurements. This approach relies on redundancy, which may lead to unacceptable increases in system weight and cost. This thesis presents an alternative, model-based approach to fault detection for a non-redundant air data system. The model-based fault detection strategy uses robust linear filtering methods to reject exogenous disturbances, e.g. wind, and provide robustness to model errors. The proposed algorithm is applied to NASA's Generic Transport Model aircraft with an air data system modeled based on manufacturer data provided by Goodrich Sensors and Integrated Systems. The fault detection filter is designed using linearized models at one flight condition. The detection performance is evaluated at a particular reference flight condition using linear analysis and nonlinear simulations. Detection performance across the flight envelope is examined, and scheduling and blending techniques used to improve detection robustness across an expanded flight regime are explored.