Motion Cueing Quality in Driving Simulation

Abstract

Driving simulators are important tools in supporting the research and development of automotive systems. The control of the simulator’s motion system is performed by the motion cueing algorithm (MCA), which converts the motion of the real vehicle to motion that fits in the constrained workspace of the simulator. Depending on the simulator and the parameter settings of the MCA, this nevertheless inevitably leads to mismatches. When perceived by the driver, these are known as incongruences. The simulator and MCA choice is a largely subjective task, for which currently little guidelines exist. This dissertation provides a novel and systematic approach to determining the motion cueing quality to guide the choice of the available simulators and MCAs. This is done by developing predictive models of subjective ratings based on objective simulator motion. This prediction method is then employed to make an estimate of the potential motion cueing that may be achieved once an MCA parameter tuning would actually be performed. Furthermore, metrics of cost and tuning are included, to provide a fairer and broader analysis, apart from the quality of the motion itself. The results show that the ratings of participants in urban, highway, and rural scenarios can be accurately predicted through low-pass filtered responses to the longitudinal and lateral specific force mismatch, as well as the yaw rate mismatch. Furthermore, an example analysis of the potential motion cueing quality, including cost and tuning metrics, is provided to lay the foundation of future motion cueing evaluation processes, and shows that including these metrics affects the total motion cueing quality. The presented results provide a method to better trade-off the various MCAs and simulators, and to reduce the necessity of on-site simulator evaluations. Combined, this may save valuable simulator time and reduce financial cost. This dissertation offers both practical guidelines and fundamental methods for im- proving the motion cueing of future driving simulation experiments. By providing a new view on motion cueing quality, the long lasting and complex problem of driving simulator motion cueing optimization is better understood.