In-vehicle comfort assessment during fore-and-aft random vibrations based on artificial neural networks (ANN)

Abstract

Driving comfort is one of the important factors for vehicle users. There is a lot of research related to comfort or discomfort in a vehicle but there is still no well-defined way to assess it accurately. In this paper, an assessment of vehicle comfort during fore-and-aft random vibrations was made based on measured and predicted r.m.s. acceleration values. Measured values of r.m.s. accelerations were obtained by laboratory testing while the predicted values of r.m.s. accelerations obtained based on the ANN (Artifical Neural Network) model. 20 male subjects participated in the study. Their different anthropometric characteristics of the body were taken into account. Based on the measured r.m.s. values of acceleration formed ANN model which has ability to predict r.m.s. acceleration values based on measured values. The obtained results showed high accuracy of the model.