Scenario discrimination of near-miss incident data

2016 Fiscal Year Final Research Report

• Project/Area Number: 26350455
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Social systems engineering/Safety system
• Research Institution: Akita Prefectural University
• Principal Investigator: Mimuro Tetsushi 秋田県立大学, システム科学技術学部, 教授 (90507112)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: ヒヤリハット / ドライブレコーダ / シナリオ / 自動判別 / 実車走行試験

Outline of Final Research Achievements

Traffic accident data and near-miss data recorded by drive recorders (DRs) are effective for active safety analyses, but automatic understandings of incident scenarios are required for the large-volume data registration / analysis.
In this research, the following three methods were developed at the viewpoint that it is not limited to the acceleration near the time of trigger. High discrimination rate of smooth / rough road was obtained by setting thresholds depending on vehicle speed for average power spectrum of vehicle vertical acceleration.in the low and high frequency ranges. By using the GPS azimuth information, high discrimination rate of four types of vehicle behaviors was obtained. Vehicle speed profiles are discriminated into three types and their characteristics were analized. We also considered utilization technology assuming future enhanced DR.

Free Research Field

自動車工学