Research On Navigation And Flight Control Methods Of Quadrotor UAV For Lifting Operation

Lifting operation using unmanned aerial vehicles(UAVs)are widely used in various scenarios.For example,when retrieving an autonomous underwater vehicle operating in sea,after the vehicle surface the sea,it can be lift and put it at the pointed ship by humans or lifting equipments.During thisprocess,the keys to success for the operation are the UAV’s trajectory tracking,the objective localizating,and accurately lifting the object.This project focuses on the operating scenarios where a quadcopter UAV lifts the ground object,and the methods of flight control,navigation,and object localization are investigated.The main research contents are as follows:(1)Based on the analysis of the flight principles of a quadcopter UAV,the kinematic and dynamic equations of the UAV are established by the description of its attitude and using the Newton-Euler formula.These equations are the basis for the design of the navigation and flight control system.(2)By analyzing using the error theory,it has been found that in the traditional Active Disturbance Rejection Control(ADRC)algorithm,Extended State Observer(ESO)estimates the states and total disturbances of the system only based on the deviations of the input signal,but can not accurately predict their future values.Therefore,the ESO is improve to enhance the estimating accuracy and a improved ESO based dual closed-loop ADRC control approach is proposed..Simulation results show that,comparing to the traditional ADRC method,when there are no disturbances,based on the developed ADRC method the system has smaller overshot and shorter response time,moreover,in the presence of disturbances,the developed method base system has smaller steady-state error.(3)In the real world,the sensors signals are time-varying and stochastic.The statistical property of noise are theoretical unknown and time-varying.However,they are assumed to be known and constant when establishing sensor models,and which can make filter algorithms divergent.To address this issue,an Improved Adaptive Kalman Filter(IAKF)-based integrated navigation algorithm is proposed to provide high-precision estimatesto the controller.The simulation results are as follows: for the IAKF algorithm,the ranges of position error of the x,y,and z axes are-2.570m-2.359 m,-2.090m-1.641 m and-0.849m-1.735 m,respectively.For the Adaptive Kalman Filter(IAKF)-based integrated navigation algorithm,the ranges of position error of the x,y,and z axes are-1.682m-1.831 m,-1.677m-0.887 m and-0.698m-1.735 m,respectively.These results indicate that the IAKF algorithm has better stability comparing to the AKF algorithm.(4)Based on the flight evaluation website of Beijing University of Aeronautics and Astronautics,hardware modules are selected to build a physical platform for flight verification.The flight data are as follows: during the process of the pole inserting the hole of the object,before the control algorithm is improved,the UAV’s x and y position trajectories exhibited significant fluctuations.When the UAV approached the target object,the movement ranges of the x and y positions of the UAV are 6.726m-7.651 m and 0.1079m-1.594 m,respectively,and the settling time is about 15 seconds.After the control algorithm is improved,the movement ranges of the x and y positions of the UAV are 6.935m-7.38 m and 0.581m-1.307 m,respectively,and the settling time is about 10 seconds.These data show that the improved control algorithms have faster response speed and smaller steady-state error.