Research On Indoor Positioning And Control Technology Of Quadrotor

The quadrotor is a kind of UAV that has been rapidly developed in recent years.It has excellent flight characteristics and good research value.Compared with fixed-wing UAVs and unmanned helicopters,the quadrotor is simple in structure and easy to maneuver.It integrates fixed-point hovering and maneuvering.It has received extensive attention in academic research and is gradually applied in the civilian and military fields.However,its indoor application and development are limited by the signal loss of GPS.Therefore,this paper focuses on the positioning and control technology of quadrotor,which is based on the indoor positioning and control technology of the quadrotor,the indoor mapping and positioning algorithm,the improved A~* algorithm and the QFT control method based on model reduction.Thus,the indoor quadrotor positioning and control platform is established.The main content of the paper is as follow:First,a nonlinear mathematical model of a quadrotor is established.According to Newton's second law and momentum moment theorem,the translational and rotational equations of quadrotor are obtained,and a nonlinear mathematical model is given.Considering the subsequent controller design,the quadrotor is divided into angular motion subsystem and linear motion subsystem.By using the small disturbance linearization method,we obtain the linear model of the two subsystems and the related state space equation.Then,the indoor positioning algorithm based on Hector-Slam is designed.Considering that the quadrotor has no odometer,it is impossible to estimate the trajectory directly through the mileage information.In this paper,the Gauss-Newton-based scan matching method is used to estimate the pose,and the bilinear interpolation method is used to solve the problem that the raster map occupation probability can not be biased.Additionally,the occupancy probability update function of the grid map is given.By constructing the quadrotor simulation model in the Gazebo environment and applying the Hector-Slam algorithm for indoor mapping,the effectiveness and feasibility of the algorithm are proved.Next,according to the characteristics of grid map and sensor used in indoor positioning algorithm,the improved 2D and 3D indoor A~* algorithms are designed.The indoor environment modeling for twodimensional A~* algorithm is introduced.Additionally,the distance measurement function,heuristic function and the data structure types of algorithm are improved.Considering future research,the improved 2D A~* algorithm is extended to the 3D,and the idea of slope conversion to plane distance is proposed.Through simulation experiments and comparisons,the efficiency and accuracy of the improved 2D A~* algorithm are proved,while the correctness and feasibility of the improved 3D A~* algorithm are also verified.Following,considering the input interference and modeling error,the quantitative feedback idea is introduced,which is combined with model-reduction method,to design the controller.The model reduction theory based on balanced truncation method is introduced,while the error estimation of the reduced order system and the original one is given.The QFT control method of SISO system is elaborated,and how to apply it to MIMO system is analyzed.The QFT controller is designed separately for the attitude control loop and the position control loop,then the simulation experiment is carried out to verify the effect of the controller and the input anti-interference performance.Finally,the indoor positioning and control platform of the quadrotor is established,and its hardware and software structure are elaborated.Based on the designed quadrotor platform,the indoor positioning experiment is carried out to verify the rationality of the whole platform and related algorithm design.