A Study on Highway Capacity and Level of Service for Geometric Design Considering Traffic Operational Conditions

(1)分析了多车道公路的服务水平及其影响因素，研究了不同几何条件和交通流条件下的实际速度和通行能力，并对影响它们的几何因素进行了广泛的分析。利用城际、城市快速路和高速公路上的车辆检测器数据，识别了坡度、重型车辆比率和天气条件等各种条件下的速度-流量关系。(2)分析了双向双车道公路的有效性测度，发现跟随者密度是最适合于难以执行超车行为的双车道公路的有效性测度。为了合理估计跟随者密度，利用车辆检测器的原始脉冲数据，研究了跟随者状态的判断标准。提出了一种同时利用车头时距和车速概率识别跟随者的方法。这种方法可以识别追随者更合理的比传统的方法，使用一个固定的车头时距值仅作为一个阈值。(3)为了在规划阶段估算包含多个信号交叉口的主干道路段的行程速度，建立了信号交叉口停车延误估算模型。该模型考虑了交叉口入口需求率、交通量、信号协调和周期长度等参数，这些参数在规划阶段可以粗略地假设。该模型在三个主干道路段进行了验证，发现其可靠性足以用于主干道规划阶段的性能检查。

(1) Analysis on the level of service on multilane highways and its influencing factors.Realized speeds and capacities under various geometric and traffic flow conditions were investigated and geometric, influencing factors to them were extensively analyzed. By using vehicle detector data on intercity, urban expressways and highways, speed-flow relationships were identified corresponding to such various conditions as gradient, heavy vehicle ratio and weather conditions.(2) Analysis on the measure of effectiveness on two-way two-lane highways.The follower density was found to be most suitable measure of effectiveness for two-lane highways where passing behavior is difficult to be performed. In order to appropriately estimate the follower density, judgment criteria of the follower condition were investigated by using raw pulse data of vehicle detectors. A methodology to probabilistically identify followers by using both of headway and speed of individual vehicles were proposed. This methodology can identify followers more reasonable than a conventional methodology which uses a fixed headway value only as a threshold.(3) Analysis on level of service on arterial streetsAn estimation model for delays caused by stops at signalized intersections was developed, in order to estimate travel speeds of arterial street sections which includes several signalized intersections on a planning stage. This model considers such parameters as demand ratios at entries of intersections, traffic, signal coordinations and cycle lengths, which can roughly be assumed at the planning stage. This model was validated at three arterial sections and found to be reliable enough for the purpose of performance check at the planning stage of arterial streets.