Interval Type-3 T-S fuzzy system for nonlinear aerodynamic modeling

Abstract

This paper presents an aircraft related application of interval type-3 (IT3) T-S fuzzy system for nonlinear aerodynamic modeling from recorded flight data. The theoretical base developed for design of IT3 T-S fuzzy system is utilized and longitudinal aerodynamic coefficients are modeled. These longitudinal aerodynamic coefficients are related with lift force, drag force and pitching moment. The flight data used for simulation belongs to ATTAS aircraft which is operating in quasi steady stall maneuver flight condition. The applicability of IT3 fuzzy system for modeling data uncertainties are demonstrated by adding the external noise in recorded data and making uncertainties associated with unvarying and missing data. The IT3 T-S fuzzy modeling method can suitably characterize the complex and nonlinear behavior of the aircraft in the presence of uncertainty and unsteadiness created due to aerodynamic / propulsive / structural interactions, separated air flow and mass properties changes etc. The performance comparison of generated IT3 T-S fuzzy aerodynamic model is also carried out with other state of the art modeling methods, the proposed fuzzy model shows good prediction capability in presence of various form of data uncertainties with associated noise. It is also noticed that as noise content in data increases, the percentage improvement in accuracy (in comparison with other fuzzy models) also appears. However, this improvement is based on nature of application and data. For clean data, the performance of IT3 fuzzy system is almost equivalent to other modeling methods, but with the increase of noise content and uncertainty in data, the performance improvement of IT3 fuzzy system enhances. Thus, performance of the IT3 fuzzy systems becomes more appealing when, uncertainty arises in the data in form of noise or missing/unvarying data points.