中文摘要：

并联机构的拓扑结构包含各支链本身的拓扑结构以及各支链相对于动/静平台布置的拓扑结构两个层面,后者对并联机构的结构、运动、动力学性能有重要影响,但研究相对较少。阐述机构结构降耦原理及其降耦设计方法,同时,提出一种基于Pro/E-ADAMS仿真运动曲线的机构运动解耦性判断方法,而无须求解机构的输入-输出位置方程。以两平移一转动3-RRR、三平移3-R//R//C并联机构为例,不改变其支链本身的拓扑结构,仅改变其在动/静平台之间的拓扑结构,设计出相应的结构降耦构型和运动解耦构型：两平移一转动机构的耦合度从k=1降为k=0,从而极易求得位置正解解析式;而原先不具有运动解耦性的三平移机构具有了部分运动解耦性,因而运动规划和控制变得容易。对一种3-R//R//C解耦构型的位置正解、运动解耦性、奇异位形、工作空间、灵巧度性能进行计算分析,发现其性能明显优于优化前的原始构型。研究成果为并联机构的拓扑结构优化及其性能改善,以及揭示两者之间的关系提供了一种有效方法。

英文摘要：

Topology structure of parallel mechanism （PM） includes two aspects, namely topology structure of branched-chain itself and of branched-chain relative to the moving/static platform. Topology structure of branched-chain relative to the moving/static platform has an important impact on structure, kinematics and dynamics performances of the PM, which is little paid attention to at present. The structure coupling-reducing principle and design methods are illustrated. At the same time, a method for identifying motion decoupling for mechanisms based on Pro/E - ADAMS simulation curve is proposed without solutions for the kinematic equations. Taking the typical two-translation and one rotation PM 3-_RRR and three-translation 3-R//R//C PM as the example, the corresponding structure coupling-reduced PM and motion decoupling PM are obtained respectively after topological structure of branched-chains relative to the moving/platform are specially designed based on the previous PMs, the results of which show that coupling degree of the two-translation and one rotation PM is reduced from k = 1 to k = 0, which leads to be easy to get its analytical forward kinematic solutions, and three-translation PM that is previously of un-motion-decoupling has derived to be with the partial motion-decoupling, which leads to be easier for motion planning and control. Comparing to forward kinematic solutions, motion-decoupling, singularity, workspace and dexterity between the typical 3-R//R//C and its motion decoupling PM, it is found that total performance of motion deeoupling PM are obviously better than ones of the previous PM. The work provides an effective method to uncover the relations between topology structure optimization and performance improvement for PMs.