A robust control approach is described for controlling the position and orientation of a six-DOF parallel link manipulator, in which one prismatic link consists of a rubber muscle actuator, so that the controlled object contains several nonlinear factors. The position of the link is also assumed to be measured by using a linear potentiometer. As a result, the associated measurement system contains some measurement noises. The control system is construted using a stochastic fuzzy controller with an integlator so as to remove a tracking error. To design the feedback gains of the stochastic fuzzy controller, a mmn-max control strategy is adopted to the systems, with unknown physical parameters. The effectiveness of the proposed method is illustrated by using some actual experiments.