Font Size: a A A

Image-based visual servoing of a VTOL UAV

Posted on:2010-12-19Degree:M.ScType:Thesis
University:Lakehead University (Canada)Candidate:Khan, Jahanzeb TariqFull Text:PDF
GTID:2448390002483498Subject:Engineering
Abstract/Summary:
Unmanned Aerial Vehicles (UAVs) are an area of significant research activity these days because of their use in dull, dangerous, and dirty tasks. Lack of human pilot paves the way for advanced and more sophisticated control algorithms that should just not control each and every aspect of flight dynamics but also accomplish the desired task at the same time. Visual Servoing is one such way of controlling the attitude and position of aircrafts by using camera as a sensor.A computer aided design (CAD) model of a Vertical Take-Off and Landing (VTOL) type UAV was developed for experimental purposes. The dynamical model of the aircraft is manipulated to facilitate the control design. The dynamics of error are modeled from first principles and then two IBVS controllers are proposed. The first IBVS controller design is based on the assumption that the target velocity is known, while the second one deals with unknown target velocity.The control strategy (in both cases) relies on the design of driving force for the translational dynamics, from which the desired orientation and thrust are extracted. Thereafter, an inner high gain feedback loop is used to generate the control torque guaranteeing the convergence of the actual attitude to the desired one.Simulation results are provided to illustrate the effectiveness of both controllers.Visual servoing is of two types: image-based visual servoing (IBVS) and position-based visual servoing (PBVS). In IBVS approach, only one camera is used to control the motion of a robot while in PBVS approach two cameras are used to guide a robot. Using PBVS approach, it is possible to obtain 3D coordinates of objects in the environment using epipolar geometry but that is not the case with IBVS approach. In IBVS approach, the depth of the object is lost due to the fact that a 3D scene is being projected on a 2D image surface. To overcome this problem, the technique of spherical projection was developed by Hamel and Mahony in 2002. In most applications of IBVS algorithms to date, one common assumption is of target being stationary which however may not be the case in certain practical applications like crime-fighting or cinematography. An attempt is made to tackle this kind of problem and an IBVS controller is designed so that the UAV can chase a moving target.
Keywords/Search Tags:Visual servoing, IBVS, Target
Related items