ディープラーニングを用いた自動車乗員身体制御モデルの検討*

抄録

Since the mass of an vehicle occupant is not negligibly small compared to the mass of the vehicle body, it is required to establish a vehicle design method considering the dynamic interaction between the occupant and the vehicle body. For that purpose, a model that can predict the behavior of the occupant when the vehicle is running is required. Therefore, in this study, we try to construct a simple mechanical model that can simulate the physical behavior of the occupant based on the data obtained from the human vibration test. In this paper, we report the results of examining a control model that combines nonlinear PD control and deep learning.