Analysis And Structural Parameter Optimization Of6-DOF Motion Simulation Platform

Compared with serial robot, Gough-Stewart platform has been widely used as motion simulators for its higher accuracy and higher load capability, etc. Compared with spherical joint, universal joint can bear more tension and rotate in larger angular range, and then universal joint has been extensively used as the passive joints of6-DOF (degree of freedom) motion simulation platforms to connect hydraulic cylinder or electric cylinder to the moving platform and fixed base. The piston and piston rod of the hydraulic cylinder and electric cylinder not only retract (or extend) along the axis of the cylinder actively, but also can rotate along the axis passively, so it is a cylinder joint, and then these6-DOF motion simulation platforms are6-UCU parallel manipulator, not the6-UPS parallel manipulator, where U stands for the universal joint, C for the cylinder joint, P for the prismatic joint, and S for the spherical joint.In order to derive good performance of the6-DOF motion simulation platform, its kinematic characteristics and dynamic characteristics are needed to be analyzed firstly, and then the good performance structural parameters are needed to be derived through optimal design.The inverse kinematic and dynamic analyses are the mechanical structure design bases for the6-DOF motion simulation platforms. In practice, the real structure of the6-DOF motion simulation platforms is using universal joints as the lower and upper passive joints of each leg, and using cylinder joint in the middle of each leg. Based on the real structure, the complete inverse kinematic modeling of the6-DOF motion simulation platforms is obtained firstly, and then the complete inverse dynamics of the6-DOF motion simulation platforms is analyzed by using the Newton-Euler method and D’Alembert principle. The movements and forces of various components of6-DOF motion simulation platforms can be derived through the complete inverse kinematics analysis and the complete inverse dynamics analysis respectively. Two special universal joint axis arrangements, which cause denominator to be zero, are derived through the kinematic analysis process and dynamic analysis process. The correctness of the complete inverse kinematic model and complete inverse dynamic model of the6-DOF motion simulation platforms is confirmed through a case study.If the6-DOF motion simulation platforms are in a singular pose in the required workspace, their kinestatic characteristics would be changed. In order to derive good performance of the6-DOF motion simulation platform, the required workspaces for the optimized mechanical structure parameters are needed with no singular posture, and then singularity analysis and singularity detection are needed in the design process. The singularities considering the impacts of the active prismatic joint and the passive universal joint are analyzed in this thesis as the universal joint axis arrangement would affect the singularities of the6-DOF motion simulation platforms. Two singular types are defined as leg singularity and actuator singularity according to the different causing reasons. The conditions caused these singularities are derived by using screw theory. Two cases of leg singularities of the6-DOF motion simulation platforms are found in this thesis, namely, when the rotational joint axis fixed on the base or fixed on the moving platform of the universal joints and the axial direction of the actuator joint of the same leg are collinear, the6-DOF motion simulation platform is in the leg singularity posture. In order to detect all the singularities of the6-DOF motion simulation platforms, singularity detection procedures are proposed correspondingly, which can directly determine whether there are singularities within a6-DOF given workspace or6-DOF reachable workspace or not. The effectiveness of the proposed singularity detection procedures are confirmed through case studies.Two commonly used performance index functions of the6-DOF motion simulation platform, condition number and manipulability index based on the kinematic traditional Jacobian matrix, are introduced in this thesis. Because the values of condition number and manipulability index based on the kinematic traditional Jacobian matrix are changed as the matrix elements in different units, then the concrete derivations of the two common homogeneous matrix formulation methods, characteristic length method and three end-effector point method, are presented. In order to measure the consistency of each actuator dynamics, a new condition number performance index function based on the generalized inertia matrix in the joint workspace is proposed. The invariant of the new condition number as the elements of the generalized inertia matrix in different units is confirmed through a case study.In order to get good performance of the6-DOF motion simulation platforms, it is needed to optimize in the design procedure. In practice, there is a multi step process to derive the final parameters of the robot not with only one step, and then multiple alternatives are needed to provide for designers after the primary design stage. The6-DOF motion simulation platform design procedure can be divided into two stages. In the first design stage, structural parameter design is needed to meet the workspace requirements of users. The6-DOF motion simulation platform design issues are divided into three categories according to different application requirements, and then the corresponding structural parameter optimization design algorithms are put forward. Because condition number and manipulability index based on the kinematic traditional Jacobian matrix are consistent with the performances of the6-DOF motion simulation platform, they are chosen as optimization objective functions. Multi-objective evolutionary algorithm NSGA-II is applied to optimize the two objective functions simultaneously to get multiple optimal solutions, and then multiple alternatives can be provided for designers in the secondary design stage. The effectiveness of the proposed structural parameter optimization design procedures is confirmed through three design case studies.