Optimal Flight Path Determination in Turbulent Air: A Modified EKF Approach

Abstract

By using the Extended Kalman Filter an accurate path following in turbulent air is performed. The procedure employs simultaneously two dierent EKFs: the rst one estimates disturbances, the second one aords to determine the necessary controls displacements for rejecting those ones. To tune the EKFs an optimization algorithm has been designed to automatically determine Process Noise Covariance and Measurement Noise Covariance matrices. The rst lter, by using instrumental measurements gathered in turbulent air, estimates wind components. The second one obtains command laws able to follow the desired ight path. To perform this task aerodynamic coecients have been modied. Such a procedure leads to a set of unknown stability and control parameters containing the required displacements of the controls. The lter estimates the new set of aircraft stability derivatives by using measurements made by the desired flight path parameters. Once the unknown stability and control derivatives have been determined, the obtained control displacements are used to perform an accurate path following in turbulent air. The obtained control laws are adaptive and they depend by either the characteristics of the disturbance or the desired ight path. The proposed algorithm, using appropriate imposed constrains, permits to tune the lter without trial and error procedure.