利用行車輔助系統探討駕駛行為之研究

Abstract

本研究透過A客運業者所提供其2017年4月至2018年3月（共1年）的里程數及肇事次數，計算其百萬公里肇事率，找出最高之肇事率調度站為光復站，以此調度站的87位駕駛為研究目標，探討這些駕駛在6個月中（2017年10月至2018年3月），統計四個項目（與前車跟車距離、壓左車道線、壓右車道線、疲勞閉眼）的異常次數之四分位數，再從各項目排名後25%區間之駕駛中取「三個以上項目皆為後25%區間」之駕駛作為本研究個案分析之目標，最後得出6位駕駛。利用輔助設備觀測此6位駕駛之駕駛行為變化，得到駕駛及四個項目的個別及整體分析。 再從此調度站的駕駛中，找出肇事率排名最高之4位駕駛，比較其四個車輛輔助項目的異常次數，探討其異常次數與肇事率之關聯性。

Based on the statistics that A Bus Company provides, this study shows that the dispatching station with the highest rate of traffic accident is Guangfu station. Aiming at the 87 drivers from this station, we study how they perform on four subjects (distance from the previous car、touching left lane line、touching right lane line、closed-eye tracking) and calculate each subject’s quartile of the frequency that abnormality happens. Then we single out the last 25% part of each subject and finally select 6 drivers who have been fallen into the last 25% range for over three subjects.

Besides, we also select four drivers in this station whose traffic accident committing rate rank the top four highest and compare the abnormality frequency those four drivers perform in the four given subjects, exploring the relevance between them and the traffic accident rate.