Lightweight Snake Robot System Design And Segmented Motion Planning Strategy Research

Due to earthquakes,landslides and other natural disasters,large areas of ruins will be formed,which will bring great difficulties to post-disaster rescue.Robot-assisted rescue has become a more efficient and safe form of rescue.Snake robot is a special mobile robot.Its slender body can penetrate into narrow spaces,has a small mass and is evenly distributed on the machine body,and can adapt to harsh environments such as loose roads.The snake robot has super-redundant degrees of freedom,making it highly flexible in movement and highly adaptable to the environment.It has the incomparable advantages of traditional mobile robots in terms of flexibility and mobility.Therefore,snake robots are more suitable for assisting rescue.In actual post-disaster rescue missions,snake robots often need to pass through a narrow space and use a head camera to collect image information on the disaster site and perform terminal operation control tasks.In order to complete the rescue mission,this dissertation developed a set of lightweight snake robot prototype and real-time control system,and studied the snake robot's backbone curve discretization method and segmented motion planning strategy.Finally,the effectiveness of the segmented motion planning strategy is verified through experiments..Aiming at the problem of lightweight snake robot prototype development and real-time control system design,firstly,this dissertation designs modular joints in structure.The joints are connected by flanges,and the joint axes are orthogonal to each other to form a snake robot.Through the customized design of mechanical structure and electrical system of robot module,the purpose of compact and lightweight structure is realized.In order to meet the demanding real-time requirements of the snake robot with super redundant degrees of freedom and the easy scalability of the control platform,this dissertation develops a snake robot real-time system based on the Linux OS with the Xenomai kernel,and adopts the Robot Operating System,ROS)to plan the non-real-time task layer.Experiments proves the reliability and capability of the real-time control system based on the Xenomai kernel.The end posture of the snake robot head output by the IMU sensor fusion is used to correct the head camera image,thereby realizing the real-time observation of the surrounding environment during the movement of the snake robot.Aiming at the limitation of the current snake robot motion planning based on the discretization of the spine curve,this dissertation appropriately improves the discretization of the spine curve.This dissertation presents the construction method and boundary constraint conditions of the general segmented ridge curve.In order to improve the discretization of the plane backbone curve,this dissertation proposes a discretization method based on the Rotation minimizing frame(RMF).Compared with the traditional discretization method based on the integral curvature and torsion,this dissertation proves the proposed method through numerical experiments.The calculation amount of the method is greatly reduced,and the discretization error is also reduced.In order to improve and reduce the discretization error of the spatial backbone curve,this dissertation proposes a discretization method based on the curve energy standard and the optimized spatial backbone curve.Based on the comparative numerical experiment results of the algorithm,this method greatly reduces the discretization error of the discretization of the spatial backbone curve compared with the traditional method.Aiming at the motion planning requirements of snake robots crossing narrow spaces when assisting in rescue and reconnaissance tasks,this dissertation studies the segmented motion planning strategy based on Bezier backbone curve.Aiming at the segmented motion planning strategy based on Bezier backbone curve,this dissertation designs a segmented spine curve that can be flexibly adjusted.Compared with the existing segmented backbone curve motion planning strategy,the discretized configuration of the backbone curve designed in this dissertation has a larger working space.The flexible adjustment of the robot's pose is achieved by controlling the deformation of the backbone curve and discretizing it.Part of the joints of the snake robot are planned to pass through the holes in the wall through the segmented spine curve,and finally the task of passing through the holes in the wall is completed and the dynamic simulation The effectiveness of the segmented motion planning strategy is verified.Aiming at the motion planning problem of the end operation control of the snake-like robot in assisting rescue tasks,this dissertation proposes a segmented motion planning strategy for the snake robot based on a flexible workspace.And the kinematics chain is divided and calculated by the flexible workspace of the robot end.In addition,in order to avoid the collision between some modules and the ground at the initial moment,this dissertation proposes an overall planning method based on Cartesian space,which solves the problem that the collision between the module and the ground.This will cause the instability of the snake robot motion.When the snake robot is stable,the accurate motion of the end of the robot's head is achieved through the kinematics segmentation solution,and simulation experiments further prove that the strategy based on flexible workspace works.Finally,the proposed strategy is compared with the existing segmented motion planning strategy in the workspace,which proves the advantages of the proposed strategy of wide applicability.In order to verify the actual effectiveness and advantages of the proposed snake robot motion planning strategy,this dissertation compares the proposed two segmentation strategies based on the Bezier back curve and the flexible workspace with the existing strategies.The effectiveness and flexibility of the proposed motion planning strategy based on Bezier backbone curve is verified through comparative experiments through narrow holes in the wall.The comparison experiment of the end motion of the snake robot proves the usefulness and accuracy of the segmented motion strategy of flexible workspace.In addition,experiments have verified the effectiveness of the head end posture correction method based on IMU sensor fusion.The motion planning strategy proposed in this dissertation satisfies the needs of the snake-like robot for deep reconnaissance and terminal operation tasks in a narrow space when assisting in rescue.