Abstract: A method to complete the setup of the surgical robot is presented, which can guarantee the dexterity, accessibility, visibility. And a virtual reality system is established to assist the surgeon to pose the robot before operation. The incision placement and the arm position are obtained by particle swarm optimization algorithm. The method not only can effectively complete setup optimization of the robot in an operation space, but also can reflect the internal difference of the operation space by dividing the operation space and assigning the subspaces with weights. Global isotropy index and cooperation ability are put forward, and their weighted average based on space division are used as objective functions to reflect the single-arm and three-arm cooperation performances of the robot respectively. Hence, the method can intuitively guide surgeons to complete the complex preoperative arrangement, taking into account the performance of the surgical robot and the internal difference of the operation space. Experiments show that compared with the method without weights, the dexterity of the three arms are all better, and cooperation capability index of the three arms increases by 25%～30%, and it can give priority to important regions of the operation space.