Set-membership fault detection under noisy environment with application to the detection of abnormal aircraft control surface positions

The paper develops a set membership detection methodology which is applied to the detection of abnormal positions of aircraft control surfaces. Robust and early detection of such abnormal positions is an important issue for early system reconfiguration and overall optimization of aircraft design. In order to improve fault sensitivity while ensuring a high level of robustness, the method combines a data-driven characterization of noise and a model-driven approach based on interval prediction. The efficiency of the proposed methodology is illustrated through simulation results obtained based on data recorded in several flight scenarios of a highly representative Aircraft benchmark.