Font Size: a A A

Research And Implementation Of Autonomous Landing For Quad-rotor UAV Based On Visual Aid

Posted on:2016-11-16Degree:MasterType:Thesis
Country:ChinaCandidate:L LiFull Text:PDF
GTID:2382330542457353Subject:Control engineering
Abstract/Summary:PDF Full Text Request
With the progress of science and technology,UAV could completed many kinds of complex tasks.Quad-rotor UAV as a multi-rotor UAV,now is the most widely used UAV.In the UAV mission,the recovery of the UAV is a very significant process.The traditional navigation equipment such as GPS and inertial navigations are limited in many cases,and the navigation system based on the vision aid is the main direction of the UAV.Through the combination of airborne vision and other sensors' information,it can realize the accurate and safe landing of UAV.In this thesis I use the Qball-x4 Quad-rotor UAV which product by Quanser company of Canada as a controlled object,using visual aids for the autonomous landing.The main research of this thesis is as follows:(1)Build Qball-X4 linear state space simulation model.Optimizing the PID parameters of Qball-X4 controller by LQR index.Then design the fuzzy self-tuning PID controllers based on LQR optimization,using Simulink simulation verify the validity of controllers,then combined with the Quanser hardware in the loop platform to realize Qball-X4 actual flight mission.Lay a good foundation for the study of autonomous landing.(2)Design the landing mark.When the concrete position of the landing mark is unknown,and the autonomous landing is accomplished by the Qball-X4 airborne visual.First use the landmark's Hu moment invariant feature complete landing landmark recognition,and the desired yaw angle can be calculated by using the Hough transform.At the same time,using STC tracking algorithm for real-time landmark tracking,the landmark center's position will be transmitted to the control input to complete the Qball-X4 autonomous landing.(3)To another situation of autonomous landing,that in the inertial navigation equipment can not provide reliable attitude data of UAV.We can use airborne vision for real-time evaluate the attitude of UAV.In the case of the landmark location was known,using airborne vision to detect the corner points of landmarks and calculate the position of the corners of the landmarks in the pixel coordinates.The attitude angles of Qball-X4 are calculated by the conversion relationship between pixel coordinates and world coordinate system,then put it as the auxiliary condition to complete autonomous landing.
Keywords/Search Tags:Quad-rotor UAV, autonomous landing, STC, fuzzy self-tuning PID control, the hardware-in-loop simulation
PDF Full Text Request
Related items