Study On Visual Tracking Control Of UAV With A Suspened Cable Load

With the rapid development of robotic technology,unmanned aerial vehicle(UAV)has been widely used in various fields in recent years due to its excellent performance.UAVs have undertaken many operations under special circumstanc es,especially in the field of civil aviation.Lots of countries and companies have invested huge amounts of money in the research of UAVs.UAVs transportation becomes an important branch in the UAV applications.The dissertation studies the transportation control by using a UAV with a suspended cable.To address the problem,we try to design an efficient attitude controller to enable the UAV with a suspended cable load flight stably.Furthermore,we proposes a visual servoing controller to drive the UAV s ystem to track a moving target.To realize the design,we first use the Euler Lagrange equation to establish the system dynamics model by considering both the suspended load and quadrotor.The whole system’s attitude controller is designed based on the model.The suspended cable load is observed in real time using a vertical downward camera,and the CamShift tracking algorithm combined with the Kalman filter is used to track the suspended load in real time.The downward camera model is established,and the two-dimensional image coordinates obtained by the CamShift algorithm are transformed into the three-dimensional physical coordinates of the suspended load with respect to the UAV base frame.In practical applications,the UAV often needs to transport a load while tracking a moving target,like a load truck.To realize the tracking,a visual tracking controller based on TLD is designed.The target information obtained by the TLD algorithm is used as reference input to control UAV flight.Because the suspended load causes serious interference to the UAV stabilization,the visual servo controller is finally combined with the UAV and suspended load position control loop.To verify the proposed controllers,a quadrotor system is built to perform several experiments.The controller for stabilizing the quadrotor with a suspended cable load,and the visual tracking controller are evaluated with a motion capture system.The experimental results illustrate that the proposed controllers perform very well.The proposed research results can be an interesting reference for other researchers in the field of robotics.