Model Predictive Control based Motion Cueing Algorithm for Driving Simulator

Thanks to the emerging integration of algorithms and simulators, recent Driving Simulators (DS) find enormous potential in applications like advanced driver-assistance devices, analysis of driver's behaviours, research and development of new vehicles and even for entertainment purposes. Driving simulators have been developed to reduce the cost of field studies, allow more flexible control over circumstances and measurements, and safely present hazardous conditions. The major challenge in a driving simulator is to reproduce realistic motions within hardware constraints. Motion Cueing Algorithm (MCA) guarantees a realistic motion perception in the simulator. However, the complex nature of the human perception system makes MCA implementation challenging. The present research aims to improve the performance of driving simulators by proposing and implementing the MCA algorithm as a control problem. The approach is realized using an actual vehicle model integrated with a detailed model of the human vestibular system, which accurately reproduces the driver's perception. These perception motion signals are compared with simulated ones. A 2-DOF stabilized platform model is used to test the results from the two proposed control strategies, Proportional Integrator and Derivative (PID) and Model Predictive Control (MPC).