Adaptive Extended Kalman Filtering for virtual sensing of longitudinal flight parameters

The works presented in this paper are a part of research studies which aim at evaluating new concepts for the control and guidance of future A/C, focusing on the augmentation of control laws' availability. The main purpose is to improve the monitoring and the consolidation of the key parameters used by these latter and by the flight envelope protections. Model-based FDD techniques, that make a global use of all or part of the sensor data available, supplemented by a simulation of a flight mechanics modeling can achieve a realtime estimation of the critical parameters and yield dissimilar signals. Filtered and consolidated information are delivered in unfaulty conditions by estimating an extended state vector including wind components, and can replace failed signals in degraded conditions, as a virtual probe would do. Accordingly, this paper describes an efficient self-adaptive EKF-based estimation scheme allowing the longitudinal flight parameters of a civil A/C to be estimated on-line. To facilitate onboard implementation, the main aerodynamic coefficients, as well as propulsion effects, are approximated by a set of surrogate models. Results are displayed to evaluate the performances of that approach in different flight conditions, including external disturbances and modeling errors. They correspond to real flight test data.