Multi-Variable Iterative Learning Identification and Its Application to Estimation of Aerodynamic Derivatives in an Aircraft Model

抄録

This paper presents a system identification technique, called iterative learning identification for multi-variable continuous-time state-space (SS) systems using iterative learning control. The transfer function (TF) parameters are regarded as functions with respect to the SS parameters that are to be identified. The relationship between the SS parameters and the response error is explicitly derived. An updated law of the SS parameters is given so as to reduce the response error. The proposed technique is applied to estimation of aerodynamic derivatives in a lateral linear model of aircraft. The effectiveness is demonstrated in numerical simulations.