Abstract: To ensure the execution accuracy of the end-point as well as the autonomy and flexibility of redundant manipulators, a precise control method based on hybrid force/position algorithm is proposed. Firstly, motion characteristics of the manipulator system are analyzed by establishing precise kinematic model. Then, the inverse dynamic can be solved by the combination of gradient descent method and fixed angle method. And the path of the manipulator end-point can be planned according to the force and position control methods by establishing the model of hybrid force/position control law. For the restricted manipulator control in complex environments, an algorithm for establishing the slave coordinate system of the redundant DOF (degree of freedom) manipulator is proposed to enable the end-effector to operate on a random curved surface. Finally, the manipulator end-point is controlled to track an arbitrary preset curve to a given point in simulation analysis and experiments. As results of multiple measurements, the errors of force accuracy are less than 2%, and the deviation values of the end-point's trajectory are less than 5% while guaranteeing the precision of force direction. Results show that the proposed hybrid force/position control method performs well in precision on a 7-DOF manipulator.