Measuring the Angle of Attack - Practical Considerations for the Development of Fault Detection Residuals

Abstract

Flight control systems depend on precise and reliable measurements of inertial and air data, which are crucial for maintaining flight safety. Air data sensors, which are exposed to harsh conditions and potential damage, are prone to common cause failures, which increases the risk of errors. Angle of attack (AOA) is particularly critical, as it impacts control laws and autopilot systems. Conventional fault detection via hardware redundancy is inadequate in cases of common cause or sequential sensor failures. Analytical redundancy methods, which integrate diverse sensor data and physical models, can potentially offer a solution. Fault detection residuals, which are computed using independent sensor signals or kinematic relationships, are central to this approach. However, false positives can occur under specific conditions. This paper focuses on angle of attack measurement and the way to check their values and trends using analytical redundancy. Various residual calculation methods are explored, with real flight data from the DLR research aircraft A320 ATRA utilized for validation. This paper analyzes the characteristics of various residuals in the selected maneuvers and conditions. It also provides practical recommendations for utilizing residuals in flight systems and discusses them.