Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : November 30, 2022 - December 02, 2022
The Tokyo University of Agriculture and Technology has a database that collects driving data in near-miss situations. The database is intended to be used f to investigate the causes of human error through accident analysis, to verify the functions of new driving support devices and to do etc., but there is a problem that the speed of the driving data has a low resolution. The purpose of this work is to study a method to estimate the vehicle speed with high resolution and high sampling rate using the acceleration and speed of the driving recorder for utilization of drive recorder speed data. Estimation was performed using the least-squares method and the Kalman filter. In each estimation, speed data was used in two different ways. The performance of these four methods was compared and evaluated. The comparison showed that the least-squares method produced the smallest error in the four estimated velocities. The estimated vehicle speeds by this method were used in the near-miss database.
