Accurate Parameter Identification of UAV by Geometric-Based, CAD Design, and Pendulum Experiment-Based Methods

In control of dynamical system, too high uncertainty of system parameters may cause fault of controller design which leads to instability. To avoid such undesired case, identification of good ap-proximation of dynamical model and its parameters is very important. In this paper, we propose methods for obtaining accurate value of each parameter for Unmanned Aerial Vehicle (UAV) with six degrees of freedom (6-DOF). We consider finding the best parameter selection method by comparing three methods; Geometric-based, computer-aids, and pendulum experiment-based method. We detail the pro-cess of experiment-based method. The test of proposed UAV is suspended by two thin parallel wires in such a way to form a torsional pendulum about the vertical axis. With RION IMU570 sensor acquiring data, the pendulum oscillations are recorded and processed in order to obtain trend- free and noise-free signals used in the final inertia estimation phase. This method can find the most accurate parameters compared to those obtained from CAD design.