Experimental study reveals accurate multi-axis vibration comfort prediction in drivers using deep learning, highlighting sensitive transmission paths for ergonomic seat design.
Ergonomics research increasingly prioritises driver comfort in complex vibration environments for safety and health. The existing evaluation methods fail to reveal the nonlinear relationship between vibration and comfort, ignoring the interpretability. This study proposes an interpretability comfort evaluation method using a hierarchical double path attention network (HDPA-Net). First, the acceleration and comfort data of the seat floor, cushion and backrest are collected through the road test. Then, the vibration transfer model and path decoupling method are utilised to identify vibration-sensitive paths. Finally, combined with a deep-learning model for comfort evaluation. The results showed that the multi-axis coupling has a significant impact on the transmission path. The HDPA-Net can capture vibration characteristics, achieving a mean absolute error of 3.3073 within a 0-100 score range, below the 10% acceptance threshold of ergonomics experts. This provides a new method for comfort evaluation and has an important guiding value for ergonomic seat design.
No takes yet. Share an insight, caveat, or question.
Bai et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: