Redundantly actuated parallel manipulator demands higher actuation coordination than the non-redundantly actuated one, thanks to its degree of actuation exceeding its degree of freedom. To improve the actuation coordination of redundantly actuated parallel manipulator, a novel model-based driving force synchronous control method with neural network was proposed. With 6PUS+UPU parallel manipulator as object, dynamic model was derived based on virtual work principle. To improve the actuation coordination of redundantly actuated parallel manipulator, a novel driving force synchronous control method was proposed on basis of force-position hybrid actuation. The synchronous control method was based on the driving force error of actuated joints and driving force adjustment was calculated by the synchronous controller. The synchronous controller was designed with neural network. What’s more, the learning law of neural network controller was derived with manipulator’s dynamic model to improve the learning efficiency. Model simulation and prototype experiment were carried out, and a performance comparison analysis with traditional force-position hybrid actuation method was made to verify the proposed control method. Comparison analysis revealed that, comparing with the traditional force-position hybrid actuation method, driving force synchronous control with neural network proposed could effectively improve the actuation coordination of redundantly actuated parallel manipulator. The results revealed that the synchronous control method reduced the driving force errors of whole manipulator by properly magnifying the control error of force actuated joint.