With the increasing demands for pilot perception-fcontrol consistency in modern high-maneuverability flight missions, the six-degree-of-freedom (6-DOF) human- centrifuge has become an important platform for high-fidelity complex flight simulation. This paper focuses on the dynamic flight simulation algorithm for the 6-DOF human-rated centrifuge. First, a kinematic model of the centrifuge is established, and the perceptual characteristics of the human vestibular system are analyzed. On this basis, a model predictive control (MPC) method is proposed, which integrates the requirements of motion simulation with the mechanism of human perception. Under the MPC framework, the method comprehensively considers the physical constraints of the platform, coordinates the tracking errors of linear and angular motions, and introduces a human vestibular model to construct a pilot sensation optimization index, thereby achieving real-time and high-fidelity simulation of flight acceleration signals. Compared with the spherical washout algorithm, the proposed method shows significant improvements in both motion tracking accuracy and perceptual consistency. This study provides an effective motion control solution for achieving high-fidelity dynamic flight simulation on a 6-DOF human-rated centrifuge.