Abstract: Abstract: Dynamics modeling plays an important role in the application of agricultural robots, which is the key to analyze the dynamic characteristics and achieve high-precision operation. This paper addressed the issue of deriving the dynamic formulation of a novel 3-DOF redundantly actuated parallel mechanism. The structure of the parallel mechanism is composed of a moving platform attached to a fixed platform through two identical PRRR kinematic chains and one PPRR chain. The parallel mechanism has two translational degrees and one rotational degree. Firstly, inverse kinematic solution of the parallel mechanism was studied by analyzing the structure property and the constraint equation; Secondly, according to the kinematics of the redundant mechanism and considering fully the impact of inertial force for each component, the inverse dynamic equation was formulated in the task space by using the Lagrangian formalism, and the driving force was optimized by utilizing the minimal 2-norm method. By investigating the contribution of each term in the dynamic model to the driving force, a simplified strategy of the dynamic model for real-time control application was proposed. Simulation and experimental results showed that the maximal value of the driving force for the parallel mechanism without actuation redundancy was 15N, but that of mechanism with actuation redundancy was 10N andthe driving force peak was reduced by 33%. It is noted that the point β=57.6460° was the singularity which should be avoided in practical application. Additionally, the maximal tracking errors for the end-effector were 0.8mm, 0.6mm and 0.068° in Y, Z and β direction, respectively. Thus the parallel mechanism system based on the proposed dynamic model can achieve good tracking performance. This research provides technology reference for further study of high precision real-time control of parallel mechanism.