Robust fault estimation using an LPV reference model: ADDSAFE benchmark case study

This paper investigates a mixed H−/H∞H−/H∞ linear parameter varying (LPV) fault estimator using an LPV reference estimator. LMIs are used to calculate the affine parameter-dependent gains of the LPV fault estimator. The design strategy is applied to a high fidelity nonlinear aircraft model provided by AIRBUS for use within the EU-FP7 project ADDSAFE, to estimate the yaw rate sensor faults in the Air Data Inertial Reference System in the presence of the parametric uncertainties. The fault detection performances in various flight conditions are evaluated using the parametric simulation.