??????·?е???????????????????

既有研究已经论证高速公路独有的运营特点，包括双能力现象。双能力现象表明，高速公路在自由运行(非排队式)条件下比在排队条件下具有更大能力。在孤立的匝道区，双能力现象的存在，决定了匝道控制在减缓高速公路和匝道交通延迟的效率。本文进一步探讨了微观仿真模型(VISSIM)是否能够在分析现象有关特征的基础上，校准复制双能力现象。本文指出，微观模拟模型用于研究匝道控制延误评价时，必须具有再现双能力现象的能力，否则就会得到错误的结论。本文发现大多数微观模拟模型，例如VISSIM，尽管使用了传统的跟弛理论，但不能自动展示双能力现象。然而，如果在VISSIM中使用一组特殊的编码参数就能再现双能力现象。利用VISSIM模型，本文分析了匝道控制和双能力现象的特征，发现匝道控制减小了双能力值的差距，尤其是采用排队政策时。由于高速公路在自由运行状态下的能力难以评佑，本文建议可用有排队状态下的能力值来代替。

Ramp Metering and the Two-Capacity in Freeway Operations

Previous studies have demonstrated a unique operational feature for freeway operations, which is often referred to as the two-capacity phenomenon. The two-capacity phenomenon suggests that freeways can obtain higher flows under freely-flowing (unqueued) conditions than under queued conditions. At an isolated ramp metering location, it is due to the existence of the two-capacity phenomenon that signifies the effectiveness of ramp metering in reducing overall delays for both the freeway and the ramp traffic. This point is first illustrated using a numerical example. The paper further explores whether microscopic simulation models (VISSIM in this case) could be calibrated to replicate the two-capacity phenomenon as well as to analyze some characteristics associated with the phenomenon. It is important to note that microscopic simulation models, when used for evaluating ramp-metering operations with a focus on delay measures, have the capability of replicating the two-capacity phenomenon. Otherwise, erroneous and invalid conclusions might result. From this study, we found that most microscopic simulation models,such as VISSIM, use traditional car-following theory, which does not automatically exhibit the two-capacity phenomenon. However, with a special coding and parameter selection, the two-capacity phenomenon can be properly replicated in VISSIM. Using the VISSIM model, we conducted analysis on the characteristics of ramp metering and the two capacity phenomenon. We found that ramp metering diminishes the difference between the two capacity values, especially when queue flush policies are adopted. While the freeway capacity under freely-flowing conditions proves to be difficult to estimate, it should have a value no less than the capacity under queue-discharge conditions.