Online Aerodynamic Model Structure Selection and Parameter Estimation for Fault Tolerant Control

TLDR

The algorithm is particularly useful for indirect fault‑tolerant flight control using model‑based adaptive control routines. This paper proposes a recursive algorithm that jointly selects aerodynamic model structure and estimates parameters to approximate time‑varying nonlinear aerodynamic behavior. The Adaptive Recursive Orthogonal Least Squares (AROLS) method extends classical ROLS, recursively updating model structure and parameters so that a new aerodynamic model can be generated after a failure. Simulation studies demonstrate that the AROLS algorithm accurately identifies aerodynamic models and improves control performance.

Abstract

This paper describes a new recursive algorithm for the approximation of time varying nonlinear aerodynamic models by means of a joint adaptive selection of the model structure and parameter estimation. This procedure is called Adaptive Recursive Orthogonal Least Squares (AROLS), and is an extension and modification of the classical Recursive Orthogonal Least Squares (ROLS). This algorithm is considered to be particularly useful for indirect fault tolerant flight control, making use of model based adaptive control routines. After the failure, a completely new aerodynamic model can be elaborated recursively with respect to structure as well as parameter values. The performance of the identification algorithm is demonstrated on some simulation data sets.