Intelligent modeling of lateral aircraft dynamic based on flight test data

Abstract

System identification technology is an important means of aircraft modeling. In order to make up for the shortcomings of traditional system identification technology, which are complicated in calculation and limited in applicability, this paper studies the aircraft intelligent system identification technology based on black and gray box modeling. The black box model which is based on the recurrent neural network establish the dynamic relationship between the input and output variables of the aircraft; while gray box model treats the motion equation and aerodynamic coefficients of the aircraft separately, assuming that the motion equation is known, and uses an extreme learning machine to establish the forward neural network model of the aerodynamic coefficients. Thus two intelligent models of the aircraft are established, the grey box model is mainly used for model prediction for a long period, and can be utilized for flight simulation, the black box model is mainly used for online model prediction, and can be applied for model predictive control.