Mendeley  

Zotero  

Aghili-Ashtiani A, Barati M. Estimating the Physical Parameters of Y-shape Octarotor Using a Combination of Recursive Least Squares and Nonlinear Kalman Filter. Journal of Iranian Association of Electrical and Electronics Engineers 2025; 22 (4)
URL: http://jiaeee.com/article-1-1753-en.html

In this paper, the physical parameters of a Y-shaped octarotor (8-propeller VTOL UAV) have been estimated using a combination of Kalman-filter-based methods and the recursive least squares (RLS) method, where the number of sensors is less than the length of the state vector and the measurements are noisy. Knowing the relatively accurate value of the parameters helps designing a suitable model-oriented controller. The estimated parameters are: the body moments of inertia (I_x, I_y, and I_z), the rotor moment of inertia (J_r), the thrust coefficient (b), and the drag coefficient (d). To make the problem closer to reality, it is assumed that: (1) Out of 12 system states, only 6 states are measurable; (2) The measurements are noisy. Therefore, 6 unknown states are estimated by suitable filter from the Kalman family, at first. Then, the estimated states are used in the parameter estimation using the RLS method. The proposed estimation operation can be performed online and during the flight of the drone. Since the parameter estimation requires data with sufficient richness, for on-the-fly data acquisition, the control loop is closed with a not-precisely-tuned PID controller. The simulation results for various scenarios confirm the success of the proposed identification plan.