號誌化路口風險分析與安全評估模式之研究

Abstract

在過去交通安全相關研究的分析結果，大多可能隱藏了肇事個體性差異，不容易據以提出具體的安全改善措施，亦不易由此瞭解肇事發生與交通工程措施間的相關性。為使路口肇事因素盡可能具體呈現，遂將四叉路口按照車輛流動特性及方位分為進入、通過與離去共18個區位。透過對肇事進行統計分析，本研究以事故型態頻率出現較高的整合分成鄰向穿越型、對向穿越型、同向左轉型與同向右轉型四個肇事種類，並於交叉口環境因子、幾何設計因子、交通管制因子與交通特性因子中謹慎評估後選擇適合之肇事因子納入研究考量，探討路口肇事風險與這些因子的相關性，據此建構號誌化四叉路口肇事風險分析與安全評估模式。另本研究亦探討機車在路口有無兩段式左轉之路口肇事型態，並藉由分析結果推測機車兩段式左轉對降低路口風險有助益。 研究結果顯示，鄰向穿越型風險模式可以平均行駛速率、車道連續性、設計路型及清道時間衡量；對向穿越型風險模式可以直行速率、照明配置、轉向(號誌)型式、儲車空間、相關標誌設置及設計路型衡量；同向左轉型風險模式則以照明配置、轉向半徑、漸近路口線型、左轉專用道、儲車空間及車道連續性衡量；同向右轉型風險模式可以平均行駛速率、槽化島設置及轉向半徑衡量。其針對所得結果與資料觀察可提出相關改善措施，例如鄰向與對向穿越型肇事，可由改善號誌燈位置、檢討其續進號誌設計或清道時間是否足夠；而根據設計路型差異所規定的轉向限制，應有清楚的相關標誌指示或限制，減少肇事風險。 透過此肇事風險分析與安全評估模式，可進行路口肇事風險的預測評估，並可針對路口改善方案做安全績效評估。此外，依照四叉路口各肇事風險評估結果，據以判定高風險之易肇事路口門檻值，以決定路口改善措施之優先順序的參考。

The conclusion of the analysis related to the traffic safety in the past may mostly come with hidden differences contributed by individual accidents. It is neither easy to present a concrete way for improvement. nor to find the correlation between the accidents and the traffic facilities. In order to make the accident factors as concrete as possible in this research, the intersections are categorized into 18 types of blocks that include entry, pass, and exit according to the characteristics and directions of the traffic flow. Through the statistical analysis of the accidents, this study divides the accidents of high frequency into four accidental categories, including adjacent passing, reverse passing, parallel left-turn, and parallel right-turn. To explore and establish the arguments for the study, factors of intersections, environment, traffic controls, and traffic characters have been taken into consideration, therefore the investigation of the relation between the accident risk of intersections and these factors helped build the model of signal intersections in risk assessment and safety assessment models. Besides, this study also investigate the accident patterns with conditions that motorcycles with and without direct left turns at intersections, and then speculation on the advantage of reducing the intersection risk by the analyzed data was conducted. According to the conclusion of the analysis, the risk model of adjacent passing makes the average driving speed measurable, including the lane continuity, the lane design, the clearance time; the risk model of reverse passing measures the straight speed, the light configuration, the traffic signal type, the left-turn space of car standby, the related signal configuration, and the lane design; the risk model of parallel left-turn is concluded to be abke to measure the light configuration, the turning radius, the asymptote lane type, the lane for a left-turn purpose, the left-turn space of car standby, and the lane continuity; the risk model of parallel right-turn can measure the average driving speed, the jug handle installation, and the turning radius. With the discussion above being complete, ways of improvements are presentable based on the finalized analysis and the observation of data. For example, the accident of adjacent and reverse passing can be reduced by modifying the position of the signal, reviewing the design of progression signal, or reestimating whether the clearance time is enough. Moreover, the limit of turning based on the difference of lane design must have clear signals or limits to lower the accident risk. Through the model of the accident risk analysis and the safety assessment, not only can it predict and assess the accident risk of the intersections, but also assess the safety of improving the intersections. Besides, the threshold values of the high frequent accident were determined according to the conclusion of accident risk assessment in the intersections, the values are references for prioritized improvement approaches.