Evaluation of intelligent electric vehicle carsickness performance considering passenger motion sensitivity

Abstract

To address the issues of individual differences, long testing periods, and high costs in evaluating vehicle motion sickness, we propose a method to assess passenger motion sickness based on vehicle motion parameters and passenger sensitivity. The short Motion Sickness Susceptibility Questionnaire (MSSQ-short) classified participants into high and low susceptibility groups. Through real vehicle tests, we collected subjective motion sickness experiences and physiological parameters along with vehicle motion parameters (longitudinal, lateral, and vertical acceleration). Using subjective Car Sickness Ratings (CSR) and single-factor variance analysis, we identified electrodermal activity as the most correlated physiological parameter. Pearson correlation analysis linked human and vehicle motion parameters, constructing a correlation matrix to analyze the impact of vehicle motion and ride duration on motion sickness. A ridge regression model was then developed to evaluate motion sickness based on vehicle motion parameters, ride duration, and passenger sensitivity. Experimental validation showed the model achieved R² = 0.85 and an accuracy rate of 84.78%, effectively mitigating the impact of individual differences in traditional testing methods, reducing costs, and providing a theoretical foundation for evaluating the motion sickness performance of intelligent electric vehicles.