Abstract: In order to reduce influence of nonlinear factors (force ripple, load disturbance and so on) on steady time of a planar High Speed/High Precision Parallel Robot, we design a novel fuzzy self tuning PID controller. When the position error is large, the controller adopts linear PI to ensure the system's smoothness and stability; when the stop command is executed, fuzzy self-tuning mechanism is applied to eliminate static error quickly. Fuzzy self-tuning mechanism adopts a "one dimension input-two dimension output" reasoning structure to avoid coupling influence in the course of gains self tuning; different gains have different fuzzy fields of error E to save tuning time and reduce the number of rules. Experiments demonstrate that Robot's steady time has been shortened significantly after the novel fuzzy self tuning PID controller is applied to the system.