| 摘要: | 一直以來，道路交通安全都是被用路人所關注，在交通事故分析或鑑定結果，經常將駕駛人行動狀態視為事故發生主因，忽略道路環境、天候條件、車輛條件等因素；實際上，道路交通事故發生時影響事故發生因素以及特性亦相當重要，對用路人而言事故嚴重程度也是重點之一，本研究希望能夠透過交通事故資料找出影響肇事程度之重要因素，提供易發生A1事故因素，提醒用路人避免陷於危險當中，並建構出事故嚴重程度預測模式。
本研究使用桃園市民國103至108年7月底交通事故資料進行分析，針對號誌化路口轉彎車輛對機車或行人事故嚴重程度進行研究，並找出嚴重程度影響因素，以二元羅吉特迴歸為研究方法對資料進行配適，探討事故在不同道路型態、天候狀況、管制條件、交通量條件及當事者條件下發生後嚴重程度。
研究結果顯示與A1事故機率成正向影響的顯著變數包含:行人專用號誌、發生星期、光線條件、駕駛視距、車種、行人穿越道寬度、機車或行人年齡、汽車駕駛性別；在不同道路型態情境下，變動可控制正向影響變數，A1事故最大機率為0.9981。
;Road traffic safety has been concerned by road users for a long time. In most traffic analyses, driver’s error action often viewed as the main factor of the accidents. These analyses are ignoring other factors such as road environment, weather conditions, vehicles conditions, etc. In fact, these factors and characteristics that affect occurrence of road accidents are also important. For road users, severity of the accident is also a key point. This study expects to use the data of road traffic accidents to identify the important factors that may affect the degree of the accidents. Further provides factors of A1 accidents, reminds road users to avoid being caught in danger, and constructs a model for predicting the severity of the accidents.
In this study, data for analysis were released by the National Police Agency of Taoyuan City′s traffic accident data from 2014 to July 2019, and studied the severity of motorcyclist and pedestrian accidents caused by turning vehicles at the signalized intersection, and found the influencing factors of the accident severity. Then use binary logistic as the research method to explore the severity of accidents after it occurs under different road types, weather conditions, traffic control conditions, traffic flow conditions and parties′ conditions.
The results show that the significant variables that have a positive effect on the probability of the A1 accident are pedestrian signal, the day of occurrence, light conditions, driving distance, vehicle type, width of the pedestrian crossing, age of the locomotive or pedestrian, and the gender of the vehicle driver. In different road types and scenarios. With changing the control variables, the maximum probability of A1 accident is 0.9981. |
