Real-time Model Predictive Control Motion Cueing using Motion System Kinematics

Traditionally, contribution of the motion system in motion cueing algorithm is carried out by a simplistic representation of a set of operational space related state variables. Neglecting the nonlinear inverse kinematics of the motion system as the fundamental part of motion cueing system is an unrealistic simplification with undisclosed consequences. In this paper, the complete motion cueing (MC) potential of the simulator in a more realistic manner is investigated by directly penalizing the general nonlinear inverse kinematics of the motion system. The developed real time model predictive control (MPC) motion cueing strategy enables the faithful reproduction of the perception of motion along with effectively constricting the motions within the limited workspace of the motion system. The partially linearized inverse kinematic model of the motion system specifically developed for the purpose of this study along with the constant coefficient inverse kinematics model at particular so-called neutral state value are employed to examine the performance of the motion cueing based on general nonlinear kinematical model. The results of conducted comparative simulations reveal that nonlinear MPC-based motion cueing algorithm enables more effectively handling the platform workspace while faithfully reproducing the realistic perception of motion with no computational encumbrance.