Aerodynamic Parameter Modeling Using TS Fuzzy Systems from Flight Data

Abstract

This paper presents the application of fuzzy logic theory for aerodynamic modeling of the fixed wing aircraft. For the demonstration purpose, yawing moment control and stability derivatives of ATTAS aircraft are extracted from the recorded flight data of nonlinear flight profile. Fuzzy-logic-based Takagi-Sugeno (TS) model is used to represent the nonlinear dynamics of the aircraft, which can approximate it into several local linear models. The method shows its benefit when it becomes very difficult to identify the equilibrium points (trim points) due to cross-coupling and nonlinear complexities associated with modern aircraft. The Gustafson and Kessel (G-K) based clustering algorithm is used for fuzzy partitioning of the input spaces. The Gaussian type membership functions are used and its parameters are generated from the training dataset. The fuzzy rules are extracted based on weighted least square algorithms and each rule corresponds to local linear model for mimicking aerodynamic (control and stability) derivatives. Fivefold cross-validation method is used to assess the adequacy of the generated fuzzy model. The parameter tracking trends and mean square error for training and testing data sets show commendable modeling capability of TS fuzzy model for extraction of aerodynamic derivatives.