Detection of abnormal aircraft control surface positions using a robust parametric test

For upcoming and future aircraft, one important challenge to tackle is the structural design optimization as it contributes to weight saving, which in turn helps to improve aircraft performances (e.g. fuel consumption, noise, range) and consequently to decrease its environmental footprint. Jamming and runaway of a control surface could lead to significant structural loads and consequently must be considered in the aircraft structural design. A runaway is an untimely (or uncontrolled) deflection of a control surface which can go until its stops if it re-mains undetected. A jamming is a control surface stuck at its current position. In this paper, a procedure for robust and early detection of such failures is presented and it is shown that it significantly contributes to the aforementioned challenges. Firstly, an appropriate parametric model of the control servo-loop is estimated, and secondly, a fault is detected by means of a suitable decision test in the parametric space. It is shown that a particular parametric direction can be identified which is sensitive to the occurrence of the investigated faults. The proposed ap-proach satisfies technical requirements in terms of false alarm, detection time and computational burden for real implementation. Experimental results with in-flight recorded data sets provided by Airbus are presented to show the efficiency of the proposed technique.