Singularity-oriented design and analysis of parallel robots

并联机器人包括通过多个铰接腿连接到基座的移动平台。由于其无可争议的优势，更高的速度，精度和刚性，并联机器人正迅速成为任何应用的标准，其中约束运动是必需的。已经有一百多种商业上可用的产品是基于并联机器人的（例如，运动模拟器，工业机器人和机床）。然而，并联机器人也有一些不可忽视的缺点。这些机器人通常有一个小的工作空间，被不受欢迎的奇异点分割。工业中用于消除奇点的传统方法基本上是将已经很小的工作空间限制为只有一个无奇点的部分。显然，这种方法加剧了并行机器人工作空间小的问题。提出的研究计划的第一个目标是在早期设计阶段消除或减少奇点。将探索的三种可能的方法之一是合成永远不会有奇点的并联机器人，无论关节限制、驱动器行程或腿长如何。另一种可能的方法是使用冗余致动器来完全（而不仅仅是部分）消除给定工作模式（逆运动学问题的给定解）的奇异性。第三种方法是设计具有降维奇点的非冗余并联机器人（例如，用曲线代替曲面）。提出的研究计划的第二个目标是研究倒钩并联机器人。这些机器人可以在不经过奇点的情况下切换装配模式（直接运动学的解）。在不知道其交点的情况下使用倒钩机器人显然是危险的。当角锥机器人进入无奇点装配模式转换路径时，平台将以完全意想不到的方式移动。

Parallel robots consist of a mobile platform connected to a base through a plurality of articulated legs. Due to their undisputable advantages of higher speed, precision, and rigidity, parallel robots are quickly becoming the standard for any application where constrained motion is required. Already, more than a hundred commercially available products are based on parallel robots (e.g., motion simulators, industrial robots, and machine tools). However, parallel robots have certain disadvantages that are not negligible. These robots usually have a small workspace that is segmented by undesirable singularities. The conventional approach used in industry to eliminate singularities is basically limiting the already small workspace to only one of the singularity-free segments. Obviously, this approach exacerbates the problem of small workspace in parallel robots. The first objective of the proposed research program is to eliminate or reduce singularities at the early design stage. One of the three possible approaches that will be explored is to synthesize parallel robots that can never have singularities whatever the joint limits, actuator strokes, or leg lengths. Another possible approach is to use redundant actuators to completely (not only partially) eliminate singularities for a given working mode (a given solution of the inverse kinematic problem). A third approach is to design non-redundant parallel robots that have singularities of reduced dimension (e.g., curves instead of surfaces). The second objective of the proposed research program is to study cuspidal parallel robots. These robots can switch an assembly mode (a solution of the direct kinematics) without passing through a singularity. It is clearly dangerous to use a cuspidal robot without knowledge of its crosspoints. If a cuspidal robot enters a singularity-free assembly-mode changing path, the platform will move in an entirely unexpected way.