Sensory-Driven Design Method for Personalizing Ride Comfort Tailored to Individual Vibration Perception

<div class="section abstract"><div class="htmlview paragraph">This study aims to develop a design method that tailors the ride comfort and design variables of vehicle components according to individual differences in vibration perception. In conventional development, variations in vibration perception have been recognized; however, quantification methods remain undeveloped, preventing designs from being adapted to individual driver perceptions. The two unresolved problems include the uniformization of vibration perception in sensory performance modeling, which predicts sensory scores from vehicle vibrations, and design approaches that focus on minimizing vehicle vibrations without considering vibration perception. First, the authors’ previous study quantified the existence of individual differences in vibration perception through sensory scores obtained from ride simulator experiments involving 24 non-expert drivers using vibrations derived from a uniform vibration perception. Hierarchical clustering identified four perception groups; however, the reduced sample sizes resulting from the classification necessitated an uncertainty evaluation in the sensory performance modeling. This study formulated personalized sensory performance models for each group using an ensemble learning method with bootstrap resampling data to enable an uncertainty assessment. These models demonstrated sufficient prediction accuracy for sensory scores through cross-validation. Second, these models were integrated into a sensory-driven design method previously proposed by the authors to maximize sensory scores. This integrated method was validated by designing a unique electronically controlled damper system that yielded personalized solutions for ride comfort and design variables tailored to the vibration perception of each group. The practicality of the method lies in the personalization of vehicle performances with the potential for enhanced tailoring in future Software Defined Vehicles.</div></div>