快速道路號誌路口變換時段下之駕駛行為研究

Abstract

變換時段係為與路口安全關係最密切之號誌時制設計單元，其傳統之設計方法係採定性公式，然而實際駕駛決策並無法如公式之理性與準確，因此此課題的發展有了隨機性與駕駛行為的考量。 本研究特針對易發生嚴程事故之怏速公路號誌路口，現場調查個別車輛在臨近路口面對黃燈時之行止決策行為，據之構建出Logit二元選擇模式，再以總體車流模式分析個別決策之總計結果。 經分析發現快速公路車輛在面對變換時段時較行駛於一般公路者傾向於採取穿越；而重型車、臨近路口，以及離路口較近的車輛，亦均有較高之傾向選擇穿越；再以車流模式作總體分析後，發現於相同設計速度或路口環境類似的的臨近路段上，相異臨近速度的車輛所需黃燈長度差異不大，至流量飽和時所需黃燈長度始有遞增的現象。 因此對於各種道路條件與交通狀況，可應用本研究這套能同時反映實際駕駛行為與車流整體特性之方法進行廣泛分析，研訂出合理的變換時段設計準則。

The change interval is a subject of signal timing, which is most relevant to safety of intersection. The design length of change interval is traditionally according to definite physical formula. However, driver decision in real world can't be as precise or reasonable as the way predicted by definite physical formula. Therefore, the development of this subject began to focus on the random property and uncertainty of driver behavior. This study chooses some intersections of the expressway, where easy to have serious accident, investigating the decision behavior of individual driver during yellow time; then, according to this, constructs a binary-logit choice model, and analyzes results of individuals' decision by traffic stream behavior model. After analyzing, it can be found that, during the yellow time, vehicles on expressway are more likely to choose crossing the intersection than those on general highway are; heavy-vehicle, the vehicle with higher speed, and the vehicle closer to stop line are also more likely to choose crossing the intersection. It can also be found that, under full capacity, the yellow time requirement of vehicles with different approach speed is similar on the identical speed design or similar approach. The yellow time requirement will increase while above jam density. Therefore, under any condition of roads and traffic, the method, which can reflect real driver behavior and property of traffic stream, in this study, can be applied to analyze and research guideline of change interval timing.