Research On Path Planning Of Manipulator Based On Dynamics Formulation

Path Planning plays an important role in manipulator control system. Its target is to avoid manipulators contacting landers or the ground when landing, minimizing manipulators’ movements and the time needed to reach expected positions. Meanwhile the axis torque safety of manipulators should be satisfied as well. The manipulator studied in this paper is a 4-DOF orthogonal flexible manipulator on the Lander. This thesis is based on the kinematic model of manipulators, focusing on the research of forward kinematics and inverse kinematics, manipulator path planning and dynamics models of manipulators.First of all, the mathematical model of manipulator was created according to Denavit-Hartenberg rules. Based on the transitive of the structure of the chain system, the position of any point on the manipulator can be calculated. The kinematical model was simplified to a 3-DOF posture calculation and 1-DOF orientation calculation. The two parts of kinematical calculation are independent. The position and attitude determination problem of this manipulator was solved based on Euclid norm, so the accuracy of the result of collide detection and that of path planning can be guaranteed.Secondly, the safety of the path planning should be certificated by the collide detection. Because of the high precision and efficiency, traditional AABB algorithm cannot satisfy the requirements. This thesis demonstrates a new method that improves the old algorithm. This new algorithm increases the speed of collide detection by reducing the number of traversal times and improving precision by using virtual sensors.Then, the dynamics formulation model of the manipulator can then be established by Lagrange scheme. The static torque of the target joint angle of the manipulator can be calculated. The target joint angle was corrected by the method of flexible compensation. The feasibility of the method of flexible compensation is verified by the measurement data of the experiment.What is more, a new algorithm, which are acquired by improving the A* algorithm, are introduced in this thesis. It is the SA* algorithm, which minimizes the number of switches of manipulator angles. The paper used A* algorithm, which replaces the Hamming distance by the Euclid distance to satisfy the optimization of manipulator traveling time. The axis torque should be less than the max torque which the axis motor can provide, and the collide should be detected during the execution of the algorithms.Lastly, a 3D motion simulation software system was developed based on MFC and Open Inventor on VS2010 for manipulator. The simulator system of 4-DOF manipulator shows the validity of the kinematics model and path planning method, and the result shows the achievements of path planning.