Data-Driven System Identification of an Aeropendulum

Abstract

In this work, we propose a refinement of the aeropendulum platform model, specifically focusing on the modeling of thrust force. We have derived a mathematical model, proposed acquisition schemes for data collection, and conducted experiments to describe the previously overlooked dynamics of the actuator. We have developed a simulation model that comprehensively captures the system’s electrical, mechanical, and aerodynamic behaviors by utilizing electrical and mechanical information. The results obtained from the simulation data demonstrate the effectiveness of the WSINDy algorithm in accurately determining the thrust coefficient and predicting the dynamics of other variables during the testing phase.