Verification and validation of a FDD system for identification of aircraft control surface jamming

Abstract

This paper presents verification and validation results of a model based fault detection and diagnosis (FDD) system for reliable detection and identification of aircraft control surface jamming at small deflections. Three main factors contributed decisively to a high reliability of the evaluated FDD system: (1) employing an accurate model of the surface-actuator dynamics in form of a linear parameter-varying model for the synthesis of the fault detection filter; (2) performing an integrated closed-loop tuning of the free parameters of the FDD system using multi-objective optimization techniques to guarantee the lack of false alarms and missed detections, as well as satisfactory detection time performance; and (3) employing numerically sound real-time signal processing techniques to perform fault identification. The verification of the FDD system has been done using standard Monte Carlo simulations, while for validation global optimization-driven worst-case search based robustness analysis has been employed. The final industrial validation has been performed on the AIRBUS actuator test bench and confirmed the satisfactory performance of the FDD system in a true industrial setting.