匝道與地面道路號誌最佳化協控模式之研究

交通壅塞情形一直是許多民眾困擾的問題，不僅造成停等次數和旅行時間的增加，車輛排放的廢氣也因此變多，使空氣污染的情形惡化。為了提升道路使用效率，壅塞問題必須透過相關交通控制措施解決。本研究以國道1號及台74線上下匝道與地面道路作為案例分析的對象，根據車輛偵測器資料，對研究範圍進行現況分析，結果指出大雅交流道和以大雅交流道為中心的周邊道路，會在某些時段發生嚴重的壅塞，如地面道路壅塞回堵至下匝道影響主線車流，及主線車輛過多導致壅塞向上游傳遞，造成上匝道之等候車隊。因此，本研究希望提出一套匝道與地面道路號誌最佳化協控模式，考量上下匝道及地面道路之車流，針對當時交通狀況求解最佳化之號誌時制，並計算儀控率，控制上下匝道車流，以提升路網績效，減緩壅塞情形。本研究結合地面道路號誌最佳化模式以及模糊邏輯匝道儀控模式，建立號誌協控模式，對研究範圍內之路網進行控制。前者以號誌最佳化軟體Balance為基礎，每300秒根據車流狀況產生新的號誌時制，並於微觀模擬軟體Vissim直接更新；後者運用模糊邏輯演算法，選定高快速公路流量、地面道路流量、高快速公路車速、匝道佔有率、高快速公路上游佔有率及其下游佔有率六個狀態變數，並建立隸屬函數及模糊邏輯規則，以求出儀控率。此儀控率利用外掛軟體VisVap，導入模擬路網進行控制，直接於模擬路網中結合兩種模式。模擬結果顯示，對於1.5倍現況流量之路網，匝道儀控模式相對無實施儀控，能提升環中路、高快速公路西向以及部分上匝道之旅行速度；號誌協控模式則有效改善地面之壅塞情形，提升旅行速度並顯著降低行車延滯。

關鍵字

模糊邏輯；匝道儀控；號誌最佳化；適應性控制；幹道號誌控制

並列摘要

Road congestion has always been a problem for many people. It causes increased travel time and more exhaust emissions from vehicles, which pollutes the air. In order to increase the efficiency of road use, congestion must be resolved through some relevant traffic control measurements. The research takes the on-ramp and off-ramp of freeway No.1 and provincial highway No.74 and the surrounding ground road as an example for case study. According to the vehicle detector data, the current traffic situation of the research scope is analyzed. The results indicate that serious congestion occurs at certain times, some caused by congestion on the ground road where traffic waves travel backwards to the off-ramps and even the freeway mainline; some caused by too many vehicles on the freeway mainline where traffic waves propagates upstream, resulting in the queue on the on-ramps. Therefore, the research proposes a model coordinating ramp metering and ground road signal control optimization, which takes the traffic flow on the off-ramps, on-ramps, and the ground road into consideration. It optimizes the signal programs depending on the current traffic situation; and calculates the metering rate to control the flow inbound and outbound the freeway, which enhances the performance of the road network and reduces congestion. The research combines the ground road signal control optimization model and the fuzzy logic ramp metering model to create a coordinated control model for the road network within the study area. The former is based on the optimization software called Balance. It generates a new signal program according to the traffic flow situation every 300 seconds and is directly updated in the simulation software Vissim; the latter uses fuzzy logic algorithm, and the selected input variables include freeway flow, ground road flow, freeway speed, ramp occupancy, freeway upstream occupancy, and freeway downstream occupancy. The membership functions and fuzzy logic rules of them are established to determine the metering rate. This metering rate is applied in Vissim by the plug-in software VisVap to directly combine two models in the simulation road network. The simulation results show that for the road network where the flow is 1.5 times of the current situation, the ramp metering control model increases travel speeds of HuanZhong Road, westbound freeway, and some of the on-ramps compared to no control. Furthermore, the coordinated control model can effectively solve the congestion problem on the ground road, increases travel speeds and significantly reduces vehicle delay.