摘要：This paper deals with an optimization approach to design a cable driven parallel robot intended for upper limb rehabilitation tasks. The cable driven parallel robots have characteristics that make them best candidate for rehabilitation exercise purposes such as large workspace, re-configurable architecture, portability and cost effectiveness. Here, both the cable tensions that are needed to move a wristband as well as the workspace need to be carefully optimized for fulfilling the prescribed operation tasks. A specific case of study is addressed in this work by referring to LA RM wire driven exercising device (LAWEX), which is applied to upper limbs exercises. To that end, a motion capture system is used to collect quantitative data on the prescribed workspace of a human upper limb. A specific optimization problem is settled up for considering combining two optimization goals, namely, the smallest robot size reaching a prescribed workspace and the minimum cable tension distributions. A sequence of optimization steps is defined using Genetic Algorithms (GAs) applied to LAWEX robot. The proposed objective function is based on a mathematical formulation of the power of a point with respect to bounding surfaces in combination with a performance index to show the distributions of the minimum cable tensions.