APPLICATION OF NEURAL NETWORKS IN ASSESSING BODY COMFORT DURING DRIVING

Abstract

In transportation, comfort is a cornerstone of vehicle quality assessment, significantly impacting drivers and passengers alike. Despite extensive research, a precise method to quantify driving comfort remains elusive. This study addresses this gap by investigating the effects of whole-body vibrations during driving across diverse speeds and road conditions, integrating measured and predicted root mean square (r.m.s.) acceleration values. Real-world road tests provided empirical r.m.s. acceleration data, capturing the dynamic vibrational environment, while an Artificial Neural Network (ANN) model was developed to predict these values with high accuracy. The research involved ten male participants, whose varied anthropometric characteristics ensured a comprehensive analysis of vibration effects. Results demonstrated the ANN model’s exceptional precision in forecasting r.m.s. acceleration, offering a robust framework for evaluating whole-body vibration and its influence on comfort. By combining practical road data with advanced predictive modeling, this study introduces an innovative approach to understanding and enhancing driving comfort. This synergy of empirical and computational methods provides fresh insights into optimizing the travel experience across diverse driving scenarios.