Kinematic Analysis And Planning Of Wheeled Robot On Spherical Surface

Spherical pressure vessels are widely used in petroleum,petrochemical and other fields.In order to ensure their safe operation,it is necessary to effectively detect the weld defects of spherical pressure vessels.The traditional manual inspection has the problems of low efficiency,high risk,and long inspection period.Therefore,it is of great practical engineering significance to study the autonomous inspection robot applied to spherical pressure vessels.This paper starts from the analysis and planning of the wheeled robot’s movement on the spherical surface.The main work and research methods are as follows:1.A kinematic modeling method is proposed for the motion characteristics of a wheeled robot on a spherical surface.By making the tangent plane and the normal plane at the center point of the robot,the speed of the robot in a three-dimensional environment is decomposed and projected in a two-dimensional coordinate plane for description,and the trajectory and speed of the robot on a spherical surface are obtained.Then the kinematics model of the robot on the sphere is established.The discretization of the kinematics model and the simulation in MATLAB verify the feasibility of the model.2.Aiming at the problem of the positioning of wheeled robots in spherical tanks with known environmental information,an improved extended Kalman filtering algorithm based on road signs is proposed.By studying the existing positioning system,a positioning system for wheeled robots based on environmental maps on a spherical surface was established.In order to realize the recursive estimation of the robot pose,the robot localization problem is transformed into a nonlinear statistical system filtering problem,and the Extended Kalman filter is studied.The state process model and observation model of the wheeled robot moving on the sphere are designed.Based on this,an improved Extended Kalman filter algorithm based on landmarks is proposed,which can realize the high precision positioning of the wheeled robot on the sphere.3.A search method based on the combination of unit decomposition method and Fleury algorithm is designed for the path planning problem of wheeled robot traversing the weld of spherical pressure vessel.The geometric feature map representation method is used to model the environment of the spherical pressure vessel in MATLAB,and the welding seam in the environmental model is decomposed into units to construct the Eulerdiagram of the welding seam of the spherical pressure vessel.The Fleury algorithm is used to find a shortest path.The shortest path achieves the best in total distance,detection coverage and repetition rate,but it has the disadvantage of more turns.4.A weld detection and search method based on genetic algorithm is proposed,which solves the problem that the Fleury algorithm finds the number of turns in the shortest path.This method combines the task requirements of the wheeled robot to detect the welding seam,designs the vertex and edge coding methods and genetic operators,and establishes the fitness function according to the evaluation index of the wheeled robot welding seam detection path planning.Compared with the shortest path obtained by the Fleury algorithm,the optimal path obtained by using the weld detection and search method based on genetic algorithm has obvious optimization in the number of turns.