Preliminary study for motion sickness reduction in autonomous vehicles: an MPC approach

This study shows a new method for evaluating the optimal speed profile for a given route. The article deals with the feasibility of autonomous driving (AD) strategies that take into account passenger comfort, focusing on the motion sickness (MS) phenomenon. In this paper a Non-linear Model Predictive Control (NMPC) approach is used to model a vehicle moving along a predefined route; the vehicle is modelled using a point mass model and is assumed to move along a spline. Literature models are used to model MS. The Model Predictive Control (MPC) used is non-linear because the model of this article is integrated in the space domain instead of a traditional integration in the time domain; this approach is shown in several papers concerning autonomous driving control. The main contribution of this study is to implement quantitatively the consideration of comfort in autonomous driving. In the literature, the articles related to comfort in AD address the problem in a qualitative way and those related to AD control techniques analyse the problem considering only the vehicle dynamics. Another contribution is to present the spatial transformation of MS models in the literature, allowing an easier implementation of these models in AD control. The results of this introductory analysis show how MS can be reduced by minimizing the increase in travel times. This technique can be used in AD or advanced driving assistance systems (ADAS) to create less nauseogenic systems or can also be used in traditional driving by advising the human driver with the best speed profile to reduce MS.