停等車疏解特性對號誌週期長度之影響

Influences of Queue Discharge Characteristics on Signal Cycle Length for Signalized Intersections

交通界經常利用分析性模式及電腦模擬以協助訂定號誌化路口之號誌時制設計。這些模式雖然在應用上有所不同，但皆根據傳統之飽和流率的觀念。此觀念假設，綠燈開始之後，疏解率會在第4部或第5部停等車進入路口後，很快地達到一穩定之最大疏解率，此最高值即所謂的飽和流率。根據飽和流率，一車道之容量可由飽和流率、綠燈時段、號誌轉換時段、損失時間及週期長度來估計。飽和流率、損失時間及容量為經常應用於時制設計之分析性模式的三個主要參數。但最近美國與台灣的現場資料皆顯示，停等車疏解率經常持續增高，並沒有明顯的穩定值。因為實際疏解特性有別於傳統觀念，目前用於設計週期長度之模式可能不可靠。本研究利用在台灣市區38個車道所蒐集之疏解特性資料，討論使用傳統飽和流率觀念時所造成之誤差，並改良現有之模式以提供時制設計之參考。

Transportation professionals have been using analytical and simulation models for timing design of signalized intersections. Although these models differ in details and their applications, they are all based on the conventional concept of saturation flow rate. This concept assumes that the queue discharge rate at a signalized intersection would quickly reach a steady maximum after the fourth or the fifth queuing vehicle enters the intersection. This steady maximum is referred to as the saturation flow. The capacity of a lane can be correspondingly estimated by the saturation flow, the green interval, the signal change interval, the lost time, and the cycle length. Saturation flow, lost time, and capacity are three major parameters commonly used in analytical models for determining the cycle lengths that satisfy a variety of timing objectives. Recent data collected in Taiwan and the United States, however, reveal that actual queue discharge rate often does not reach a steady maximum. Instead, the discharge rate tends to rise continually long after the green onset. This discrepancy between the actual queue discharge characteristics and the traditional concept of queue discharge casts doubt about the reliability of the existing models. Based on the queue discharge characteristics observed in 38 lanes in Taiwan, this paper discusses the errors of using the conventional concept of saturation flow in modeling queue discharge and proposes improved models for signal timing design.