Recent Trends in Aerodynamic Parameter Estimation: a Review of Different Methods along with Their Advantages and Disadvantages

Abstract

This paper aims to provide a consolidated account of different recent research works on aerodynamic parameter estimation. Attempt has been made to succinctly review the work done by different authors, qualitatively compare their results, and summarize their findings. Aerodynamic parameter estimation is of great interest to researchers for past few decades. Several offline and recursive parameter estimation methods have been recorded in literature. Multiple researchers have used offline method such as filter error method for aerodynamic parameter estimation in presence of turbulence. Significant amount of work has been done using artificial and recurrent neural network methods to estimate the aerodynamic derivatives from flight data. In online parameter estimation, the stability and control derivatives are computed in real time from flight data measured through onboard sensors. In such cases different filters such as extended Kalman filter, unscented Kalman filter, adaptive unscented Kalman filter have been employed by several researchers to remove the noise and bias present in the measured flight data. The paper creates an understanding of all these methods by discussing their applications in aerodynamic parameter estimation as presented in recent research works. It adeptly unifies the recent developments on the topic from different papers, and coherently argues the advantages and shortcomings of them.