Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : September 11, 2022 - September 14, 2022
This research focuses on the feature of human dynamic motion during the use of seated personal mobility vehicles (PMV). Ten motion conditions were set up from different speed, direction, and motive methods. Two pressure sensors were used to measure the pressure and changes of center of pressure in back and hip with different speeds and directions under the autonomous conditions and manual conditions. To find out the feature of in different conditions, a deep learning program based on the ResNet were used to classify and plan to extract the feature map to show the feature of human dynamic motion. While, the poor classification result obtained is not so good because the vehicle motive in relative low speed both autonomous and manual conditions, displacement of center of pressure may not show significant difference. For this point T-test is used, and shows that both autonomous and manual conditions average displacement of centers of pressure after the vehicle braked 1 second are similar while in the beginning of vehicle starting moving are different.
