Research On Air-ground Collaborative Mapping Method Using Tethered UAV

To satisfy the requirements of radiation environment perception and detection for nuclear emergency disposal tasks,this paper aims at investigating the air-ground heterogeneous robot systems composed of tethered unmanned aerial vehicle(UAV)and unmanned ground vehicle(UGV).UAV can carry out a wide range of environmental sensing with small payload and short endurance time,and UGV have large payloads and strong endurance time but narrow sensing field and weak maneuverability.They complement each other and cooperate to accomplish the task of integral environmental perception and autonomous motion planning during the operation of the nuclear robot.The key research contents of this paper are the relative pose estimation,collaborative control and collaborative mapping of heterogeneous robot systems.The specific contents are as follows:(1)The method based on April Tag target positioning and tether position calculation and positioning was proposed to solve the problem that the relative positioning error is large for the air-ground heterogeneous robot with single sensor in GPS rejection environment.The tethered UAV accurately locates the tagged UGV by the April Tag target positioning method.To solve the problem of failed positioning caused by tag beyond the camera’s field of view,this paper researched the tethered UAV tether positioning model based on range-azimuth-pitch angle,and calculated the position of UAV in the coordinate system of UGV.Furthermore,a vision and tether fusion localization algorithm based on visual failure detection and time confidence weighting strategy was designed to improve the accuracy and robustness of the tethered UAV positioning to UGV.The experimental results show that the RMSE of the relative position of UGV and the tethered UAV in the proposed method is 0.064 m,while the April Tag target positioning and tether positioning are 0.099 m and 0.107 m,respectively.(2)Aiming at the disturbance of the tether tension on the tethered UAV during flight,a kinematic and dynamic model of the tethered UAV was established based on Newton Euler equation,and the influence of the tether tension on the position and attitude stability of the tethered UAV was analyzed.Combined with ADRC and PID control technology,a collaborative controller of the tethered UAV and UGV posture was designed.ESO and NLSEFCL were introduced into the attitude controller to estimate and eliminate the disturbance of tether tension on the tethered UAV in real time,so as to realize the stable control of the tethered UAV following the UGV.The experimental results show that,compared with the cascade PID controller,when the yaw adjustment time of the tethered UAV following the UGV increases by 4.44%,the adjustment time of the controller designed in this paper to control the UAV along the X and Y axes decreases by 26.48% and 32.40%,respectively.The maximum following error along the X axis,the Y axis and the yaw angle was reduced by68.61%,80.12% and 14.29%,respectively.(3)Aiming at the problems of weak sensing ability of single robot and incomplete information of environmental map,a local 3D point cloud height layering matching map construction algorithm is proposed.According to the structure size and motion space characteristics of the air-ground heterogeneous robot,the local point cloud maps of the air and ground are converted into layered height maps,and the layered height maps of the air and ground are matched based on the principle of 3D-NDT to solve the problems of large storage space occupied by the original point cloud map and large calculation amount of air-ground map matching.The experimental results show that compared with the original point cloud map matching based on 3D-NDT principle,the speed of the proposed algorithm is increased by98.81% under the matching accuracy is reduced by 6.82%,and the number of point clouds is reduced by 85.94%.Finally,the effectiveness of the proposed cooperation algorithm is verified by systematic experiments,which lays a foundation for the autonomous motion planning of the robot in the nuclear radiation environment.