Abstract: Orthogonal experimental design is used to synthesize the accuracy of robotic mechanism, which not only requires less calculating time, but also discovers the whole error distribution. By using this simulating method, there is no need to manufacture the physical prototype of robot. On the other hand this novel method can reduce the work of synthesizing the accuracy of the robot and simplify the process. The historical literature in this respect is concluded, and the error model of 3R robot is set up. The error level of every factor is designed by using an orthogonal table. Every group of the testing experiments is analyzed. The error distribution map of the robot under a certain task with a complex surface and the significances of every kinematic factor are proposed. The significances at current configuration of all error factors can be easily calculated by using orthogonal experimental designing, which is helpful to design the accuracy and calibrating the kinematic parameters of the robot.

Key words: Orthogonal experimental design, Positioning error, Robot accuracy, Significance