A Study on Modeling Comfort Satisfaction of Automotive Seats Based on Human Factors Engineering and Machine Learning

This study focuses on the critical impact of automotive seat comfort on user experience. Addressing existing issues in comfort evaluation systems—such as fragmented subjective indicators and the disconnect between subjective and objective parameters—we propose a hybrid modeling approach integrating ”factor analysis + XGBoost” to establish an analytical framework mapping subjective-objective parameter relationship. Subjective evaluations (using a 7-point Likert scale with 35 detailed metrics) and objective parameters (e.g., pressure distribution, joint angles) were collected from participant samples. A factor analysis model was constructed and subsequently optimized using XGBoost. The results demonstrate the successful identification of four core comfort factors, with the model effectively validating their significance. This approach provides a novel methodology for automotive seat design, bridging the gap between empirical data and user perception.