Optimization-based tuning of LPV fault detection filters for civil transport aircraft

Abstract

In this paper we propose a two steps optimal synthesis approach of robust fault detection filters for the model based diagnosis of sensor faults for an augmented civil aircraft. In the first step a direct analytic synthesis of a linear parameter varying (LPV) fault detection filter is performed for the open-loop aircraft using an extension of the nullspace based synthesis method to LPV systems. The employed scheduling parameters are either directly measurable flight conditions (altitude and speed) or aircraft parameters estimated in real time (mass and center of gravity position). In the second step, a multi-objective optimization problem is solved for the optimal tuning of the LPV detector parameters to ensure satisfactory fault detection performance for the augmented nonlinear closed-loop aircraft. Worst-case global search has been employed to assess the robustness of the fault detection system in the presence of aerodynamics uncertainties and estimation errors in the aircraft parameters. The main appeal of the proposed approach is its applicability to any robustly stable closed-loop aircraft simulation model, for which the open-loop aircraft model is trimmable and linearizable. An application of the proposed method is presented for the detection of failures in the angle-of-attack sensor.