應用存活分析法於路口直接左轉機車事故之研究;A Study on Direct Left Turn Motorcycle Accident at Intersection Using Survival Analysis Method

NCUIR > Engineering College > Graduate Institute of Civil Engineering > Electronic Thesis & Dissertation > Item 987654321/87147

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/87147

Title:	應用存活分析法於路口直接左轉機車事故之研究;A Study on Direct Left Turn Motorcycle Accident at Intersection Using Survival Analysis Method
Authors:	沈譽澄;Shen, Yu-Cheng
Contributors:	土木工程學系
Keywords:	直接左轉機車事故;存活理論;Cox等比例風險模式;交通安全評估;Direct left turn motorcycle accident;Survival theory;Cox proportional hazard model;Traffic safety evaluation
Date:	2021-08-26
Issue Date:	2021-12-07 15:08:43 (UTC+8)
Publisher:	國立中央大學
Abstract:	台灣平均每年機車事故佔整體事故約六成左右，路口是交通流匯集的重要地點，各路口之安全性影響用路人甚鉅，有關用路人行車安全之負面影響，以及時間與財務上之損失，更是不容忽略的課題。?能從左轉機車著手改善，對於減少路口事故具有相當程?的助?。本研究將路口直接左轉機車事故間隔時間作為存活分析之觀測時間，並利用存活理論中之Cox等比例風險模式應用於事故影響因素進行分析，影響因素分為道路、環境、車流、交通管制、駕駛人以及其他特性，經由模式檢定找出顯著變數，再透過相對風險分析以及中位存活時間，衡量解釋變數的改變可為存活時間增加的實質效益。研究結果顯示，經模式檢定之顯著變數，路口寬度愈小、相對濕度愈大、性別為男性、騎乘普通輕型機車以及路口事故發生在都會區，均會縮短路口直接左轉機車事故之存活時間；反之，路口設有中央分隔設施、內側車道繪製左轉指向線、設有左轉保護時相、路面乾燥，則會延長路口直接左轉機車事故之存活時間。此外，在多因子相對風險分析結果中，發現同時調整內側車道標線及轉向號誌對於延長存活時間效果最佳。案例路口情境分析中，比較了各路口相對風險變化值，評估三種路口型態安全特性，可作為降低直接左轉機車事故風險程度之參考依據。;Riders accidents in Taiwan account for 60% of all accidents on average each year. Intersection where traffic flows are important places. The safety of each intersection has a huge impact on passerby. The negative impact on the safety of passerby and the loss of time as well as finance are issues that cannot be ignored. If improvements can be made from left-turning riders, it will help to reduce accidents to a considerable extent. In this study, the time interval between incidents of directly left-turning riders at the intersection are seen as the observation time for survival analysis, and the Cox proportional hazard model in the survival theory is used to analyze the impact factors of the accident. Influenced factors are divided into roads, environment, traffic flow, traffic control, drivers, and other characteristics. Significant variables are figured out through pattern testing, relative risk analysis, and median survival time. Measuring the substantial benefit of change in explanatory variables can increase survival time. The results show that the significant variables verified by the model, the smaller of the intersection width, the greater of the relative humidity, the male, the riding of an ordinary light motorcycle, and the occurrence of the intersection accident in metropolitan area, will reduce the survival time of direct left turn motorcycle accident at the intersection. On the contrary, if the intersection is equipped with a central separation facility, the inner lane is drawn with a left turn directional line, there is a left-turn protection phase, and the road is dry, will extend the survival time of direct left turn motorcycle accident at the intersection. In addition, in the results of multi-factor analysis. it was found that adjusting the inside lane markings and turn signs at the same time, have the best effect to extend survival time. In the case of intersection scenario analysis, the relative risk changing value of each intersection is compared, and the safety characteristics of the types of intersections are evaluated. It can be used as reference for reducing the risk of directly left-turning riders accident.
Appears in Collections:	[Graduate Institute of Civil Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	109	View/Open

社群 sharing

Loading...