在城市路网中，交叉口拥堵是造成城市交通拥堵的主要原因。我国交通部门目前解决交叉口拥堵的主要措施有：加强交叉口道路基础建设、提高车道控制水平和优化信号配时方法。但是，部分措施往往存在限制，例如：加强现有交叉口道路基础建设因资金、噪音和施工时间等方面的约束，难以快速的对原有道路进行改造；在车道管理方面，左转弯待转区受限于道路渠化空间，潮汐车道虽可以改善当前道路的交通拥堵，但也易造成对向车道的交通拥堵。本文研究的问题是在基于现有道路资源上，如何在空间和时间维度上挖掘道路潜力、降低车均延误、提高交叉口通行能力。相位交换排序策略是高效利用车道进道口的一种车道管理方法，国内外诸多学者主要研究其渠化方式和信号配时方法，但是较少研究在仿真阶段考虑到不确定性驾驶行为。而元胞自动机的优势在于可以描述不确定性的驾驶行为并仿真。因此，本文以相位交换排序策略的车道与信号控制方法为出发点，对这一策略的配时方案优化，并在仿真阶段用元胞自动机描述车辆不确定性的驾驶行为，构建车辆总延误最小的优化模型，并用Anylogic搭建增强现实仿真平台，更形象地展示该策略的运行过程。本文首先介绍相位交换排序策略的车道和信号控制方法，并且阐述主、预双信号协调控制过程。然后，对该系统中车辆分别驶进排队区、待行区和驶离的过程进行分析，得到将延误最小化作为目标函数的配时优化模型，并且用遗传算法对其进行求解。其次，以机动车驾驶员在交叉口中的启动行为、加速行为、随机减速行为、随机换道行为、随机分流行为和转向行为作为元胞自动机的演化规则，建立元胞自动机优化模型。接着，将荷清路—清华东路交叉口和双清路—清华东路交叉口作为案例，用元胞自动机优化模型对这两个交叉口分析并仿真，得出结论：相较于采用传统信号控制，若交叉口采用相位交换排序策略控制，当道路饱和度达到0.9后，荷清路—清华东路交叉口中最拥堵的东进口左转及右转车道可分别降低26.96%和29.41%的车均延误；清华东路和双清路在采用相位交换排序策略后，车辆排队长度可分别降低约22.5%和27.8%。相位交换排序策略是较为复杂的一种车道控制方法，普通民众难以对其运行过程完全理解，因此本文最后搭建增强现实仿真平台，用最直观的方式呈现相位交换排序策略。
The traffic congestion of intersections in the urban road network is the main cause for urban traffic congestion. The main way to solve the traffic congestion are to strengthen road infrastructure construction at intersections, proving lane control and optimizing signal timing. However, it’s difficult to renovate the existing roads on a large scale when strengthening the construction of the existing road infrastructure at intersections due to factors such as capital, noise, and construction. In terms of lane management, the left turn waiting area is limited by the road space. Although the tidal lane can improve the current road traffic congestion, it is also easy to cause traffic congestion in the opposite lane. Thus, based on the existing resources, it is more important to propose an efficient road management method in both space and time dimensions and to tap the potential of existing roads to reduce the average vehicle delay and improve the intersection capacity.The phase swap sorting strategy is a lane management method for maximizing the utilization rate of lane entrances in terms of time and space. Many scholars in China and the rest of the world have demonstrated the effectiveness and reliability of this strategy. However, few of these studies have included drivers’ driving characteristics in the scope of studies as an uncertain factor. Based on the previous studies on the phase swap sorting strategy, this paper establishes an optimized cellular automaton model by considering drivers’ behaviors as the evolution rule and the strategy above-mentioned as the lane control method.First, this paper introduces the lane control method and signal control process for the phase swap sorting strategy and illustrates the coordinated control process of main and preliminary traffic lights. Then, the process during which a vehicle crosses an intersection in the system, i.e. entering the intersection, the queuing area and the waiting area, and then leaving the intersection, is analyzed. The minimization of average vehicle delay is used as the timing optimization method for the target function. Then, this paper adopts the driver’s behaviors of starting, acceleration, random deceleration, lane-changing, diverging and turning in a T-intersection as the evolution rules of cellular automata. A cellular automata model suitable for the phase swap sorting strategy is proposed. Finally, this paper analyzes and simulates the T-shaped intersection of Tsinghua East Road and Heqing Road, as the object of the case study, by using the cellular automata model. The following conclusions are drawn: the traffic saturation of this T-shaped intersection reaches 0.9 during rush hours; if the phase swap sorting strategy is adopted, the vehicle delay of the left- and right-turn lanes for the east entrance can be reduced by 26.96% and 29.41% and the queue length on Tsinghua East Road and Shuangqing Road can decrease by 22.5% and 27.8%. Finally, phase sway strategy is a complex lane control method, which is difficult for ordinary people to fully understand its. Therefore, this paper builds an augmented reality simulation platform to present the phase swap strategy in the most intuitive way.
