Research On Indoor Navigation And Control Of Quadrotor Based On UWB Positioning Technology

With the increasing demand for the application of UAV in indoor scene,more attention has been paid to the research of indoor navigation and flight control of UAV at home and abroad.Usually,the indoor environment is complex and changeable,thus accurate and reliable indoor navigation and flight control algorithms are the key to guarantee UAV to complete the scheduled tasks.With the background of indoor navigation and control of small quadrotor UAV,this paper proposes a method of indoor navigation and positioning based on multi-sensor fusion technology,and the quadrotor is used as a platform for modeling and design of controllers.The main research contents of the paper are as follows:Firstly,this paper chooses the quadrotor as the experimental platform,and builds a hardware platform used in indoor navigation and flight control,and then introduces the main functions of the software system.Secondly,the forces and torques acting on the quadrotor are analyzed,and then a nonlinear mathematical model of the quadrotor is established using Newton-Euler method,which will be applied in the design of the filters and controllers.Considering the need of Engineering implementation,the control efficiency model and the transfer function of the propulsor model of the quadrotor are also established.In addition,the model parameters such as the aircraft's moments of inertia and the lift coefficient of the propeller are determined by means of experiment and theoretical calculation.Thirdly,a scheme of indoor navigation and positioning system based on UWB location is proposed,which contains attitude estimation and position-velocity estimation mainly,and the EKF is designed to fuse the data of various sensors respectively.For attitude estimation problem,the EKF algorithm based on error quaternion is adopted.The effectiveness of attitude estimation algorithm is tested by experiments.For the problem of position and velocity estimation,a multi-rate Kalman filter is used to process multiple sensors with different sampling rates,and the experimental results show that the proposed algorithm can obtain an estimate with better accuracy than a single sensor.Finally,it is proved that the nonlinear mathematical model of quadrotor is differentially flat,on the basis of which,trajectory tracking controllers are designed for a quadrotor by using the nonlinear geometric control method.Aiming at the under-actuated characteristics of the quadrotor,the trajectory tracking problem is divided into two flight modes,namely an attitude controlled mode and a position controlled mode,and the controllers are designed respectively.The validity of the attitude controller is verified by simulation experiments,and quadrotor UAV can achieve aggressive maneuver flight.