壅塞交織路段之駕駛行為分析與最佳化之匝道儀控模式

Abstract

高速公路入口匝道之車流管制理論應用自1960年代開始，已經具有數十年之實際執行經驗，並獲有一定成效，包括交通壅塞之紓解與旅運時間之節省等。本文即針對高速公路分離式連接集散道路 (Separated connecting collector- distributor roadway) 單一交流道路段分析其車流特性與研究建立最佳匝道儀控模式。其影響匝道儀控成敗因素包括有車道變換行為(Lane-changing behavior)、車間距離 (Headway)、車流量、車流速率、車流密度、交織率 (Weaving rates)、主線下游車流狀況及入口等待車輛長度等。本研究之交織路段容量預估模式係應用可接受間距 (Gap acceptance) 透過Renewal Processes尋求臨界間距 (Critical gap) 及最佳線性規劃原則等建立動態容量預估模式並作為最佳化匝道儀控容量控制之依據。匝道儀控模式最佳號誌時制之建立，係假設於匝道儀控期間任何時間變化之交通需求 (Traffic demand) 是處在均一壅塞情況下，應用需求-容量模式 (Demand-capacity model) 於本研究之分離式連接集散道路交織路段進行模擬與實證。經應用現場實際調查之交通資料加以分析得知本研究之行車特性係包括100%之車道變換行為及49.29%之交織行為 (Weaving activities) 下進行，所得最佳匝道儀控率 (Metering rates) 及號誌時制，能獲得高速公路主線車流量與行車速率提升及總等候長度最少之效果，並可紓解已壅塞之車流。

This dissertation examines the conditions for which ramp metering can be beneficial to the overall system in terms of travel time savings for a simple traffic corridor that consists of a freeway and a set of parallel arterial connected by separated connecting collector-distributor (C-D) roadway and entrance ramps and/or loop-ramps. The purpose of this study focused on the traffic flow characteristics and capacity analysis, and ramp metering control for the separated connecting C-D roadway and freeway on freeway systems. As the performance of ramp metering control is concerned, it depends on various factors, such as the headway, traffic flow rate, density, speed, and flow ratios between each lane, and downstream traffic conditions, etc., and the key factor to the capacity on the interchange systems, is the lane changing activity of traffic operation in freeway weaving sections. This study developed a model to estimate the weaving capacity for the separated connecting C-D roadway on freeway systems based on the renewal processes of gap acceptance and linear optimization. Ramp metering has been emerged as an effective freeway control measure to ensure efficient freeway operations. It focuses on analyzing dynamic and control relationships to arrive at general analytical results regarding optimal metering policies. The analysis assumes that time-varying traffic demands are uniformly congested throughout the control period. The control model is using demand-capacity model to produce the optimal metering rates and the minimizing total queue length. This study investigates the development macroscopic equation models of ramp metering control signal timing on the separated connecting C-D roadway and freeway systems. It focuses on field data based on these driver behaviors including 50.71% of lane-changing and 49.29% of weaving activities to analyze. The ramp metering control is capable of optimizing mainline traffic by providing metering rates for accesses within the control segments. These results of this study including the ramp metering can increase the freeway mainline traffic flows and speeds, reach minimizing total queue length, and prevent a traffic breakdown by adjusting the metering rate.