Research On Key Technology Of Flip-type Obstacle Negotiation Robot That Can Throw

The urban environment will become a center stage for future combat operations.Confronting forces that use the asymmetric nature of urban areas to their advantage poses a real challenge to any military planners.Timely reconnaissance,surveillance and target acquisition(RSTA)capabilities are among the most important issues in urban combat environments.Aiming at key technology of flip-type robot,the optimization design of the structure is carried out to improve adaptability and stability of the robot when climbing stairs.The anti-drop ability is improved thourgh optimized design,which aims to deploy by casting.Local autonomous mobile system based on monocular vision to reduce the robot's operational difficulty and enhance the practical application of combat effectiveness is established.This paper includes the following works:The structure optimal design of the flip-type robot is performed.A two-dimensional mathematical model of the robot is established.The kinematic constraints on the flip mechanism are derived to prevent undesired interferences among stairs,wheels and main body during climbing stairs.The objective function is proposed according to the traction demand of the robot during stair-climbing motion for the first time and the value of the objective function is calculated though kinetic analysis.The Taguchi method is using as the optimization tool because of its simplicity and cost-effectiveness both in formulating an objective function and in satisfying multiple constraints simultaneously.The design parameters of the robot which can make the robot successfully climb common stairs were obtained.The performance of the robot under the optimal parameters is verified through simulations.The anti-dropping design of the stair-climbing robot with flip mechanism is performed.The drop model of four-wheeled mobile robot is established,and the response of the model under shock impact is analyzed.The relationships between the maximum compression displacements of the wheels,the maximum acceleration of the robot and the drop height of the robot,the natural frequency of the impact response are obtained.The relationship between maximum stress,maximum deformation of the shell and the thickness of the shell,the size of the load,and the maximum acceleration of the robot are obtained.e The simply supported beam is used to imitate the stress and deformation of shell under the dropping acceleration.The relationship between maximum stress,maximum deformation,thickness of shell,and the weight of load under fixed acceleration of the robot is obtained.According to the relationships of the parameter,the optimal design parameters of the anti-dropping capability the robot are chosen.A finite element model was established with ABAQUS and a free fall from the height of 6 meters to the rigid ground was tested.The results show that the optimized structure can survive from the impact to rigid ground from a free fall at height of 6 m.Local autonomous mobile system based on monocular vision is established.The image captured by monocular camera is processed to identify the scene around the robot and the position of the robot.An improved clustering segmentation algorithm to refine the SLIC segmentation procedure to better adhere to the boundaries in image is presented in this paper.An improved method of region merging based on DBSCAN is proposed to perform region merging of the obtained superpixels.The unobstructed reagion is estimated through the segmentation result of the image,and the three-dimensional data information of the unobstructed region is solved in conjunction with the pose information of the robot.In order to support achieve of the locally autonomous movement of the robot.A prototype of the second-generation flip-type mobile robot is developed.Experiments of stair climbing is carried out under different conditions of diameters of wheels,lengths of suppouting arms,turnover speeds of wheels to verify the effectiveness of structural optimization.A collapsible support arm was developed,the unford experiment and the ground-reversal support experiment of the collapsible support arm was conducted as a support for the deployment method of casting.The investigation load was added to the robot,and the investigation of the effects of the ground movement,grass movement,indoor maneuvering,and stair climbing was performed.A variety of typical indoor scene environment images are processed to obtain unobstructed areas under the scene and compared with the manual observation results,which provides support for the robot's local autonomous movement.