Monocular Camera Based Trajectory Tracking of 3-Dof Helicopter

Shuailing Li,Hao Liu,Zongying Shi,Yisheng Zhong
DOI: https://doi.org/10.1002/asjc.857
IF: 2.4
2014-01-01
Asian Journal of Control
Abstract:The vision based flight control problems of unmanned micro aerial vehicles (MAVs) have attracted much attention in recent years. This paper presents a new solution to the trajectory tracking problem of a three-degrees-of-freedom (3-DOF) helicopter by utilizing only an onboard monocular camera without any artificial marker. First, the parallel tracking and mapping (PTAM) algorithm, which is a famous solution to the visual simultaneous localization and mapping (vSLAM) problem, is employed to estimate the attitude angles of the 3-DOF helicopter. Then the calibration puzzle of the mapping between the onboard camera and helicopter coordinate systems is turned into an optimization problem. A robust cost function is applied to get the accurate estimate of the mapping. Finally, for the purpose of alleviating the influences of nonlinearity and coupling between channels, the feedback linearization (FL) and the linear quadratic regulation (LQR) techniques are employed to design the controller. Experimental results show that the proposed method can ensure that the helicopter hovers without drift and has good tracking performance.
What problem does this paper attempt to address?