本文提出了一个基于车联网（CV）环境的交通主干道信号灯协同控制的模型及其求解方法。首先， 我们将信号优化与协同控制问题归结为一个混合整数非线性规划（MINLP）。通过考虑单个车辆轨迹的最小油耗和行程时间求解最优相位持续时间和相位差。由于模型的复杂性，我们将问题分解为两个层次：使用动态规划（DP）优化相位持续时间的交叉路口层，以及用于优化所有交叉路口相位差的干道交通层。同时，我们开发了一种基于预测的方法以求解上述双层优化模型。在各种场景下，我们利用交通模拟对模型进行了测试。与传统的感应信号灯协同控制相比，求解MINLP和双层优化模型生成的信号时序规划可以合理地提升交通信号灯控制性能和路网的服务水平。在高密度的交通流量的环境中并考虑不同车辆类型时，与感应信号灯相比，上述双层优化模型的求解法使路网总成本降低了3.8%，MINLP的应用使系统总成本降低了5.9%。这也表明对于交通密度相对较高的干道交通来说，本文提出的协同控制方案效果显著。而仿真结果也表明对于同时拥有主要道路和次要道路的交叉路口而言，面向主要道路进行的协同控制对次要道路上的车辆几乎没有影响。

This study presents a connected vehicles (CVs)-based traffic signal optimization framework for a coordinated arterial corridor. The signal optimization and coordination problem are first formulated in a centralized scheme as a mixed-integer nonlinear program (MINLP). The optimal phase durations and offsets are solved together by minimizing fuel consumption and travel time considering an individual vehicle's trajectories. Due to the complexity of the model, we decompose the problem into two levels: an intersection level to optimize phase durations using dynamic programming (DP), and a corridor level to optimize the offsets of all intersections. In order to solve the two-level model, a prediction-based solution technique is developed. The proposed models are tested using traffic simulation under various scenarios. Compared with the traditional actuated signal timing and coordination plan, the signal timing plans generated by solving the MINLP and the two-level model can reasonably improve the signal control performance. When considering varies vehicle types under high demand levels, the proposed two-level model reduced the total system cost by 3.8% comparing to baseline actuated plan. MINLP reduced the system cost by 5.9%. It also suggested that coordination scheme was beneficial to corridors with relatively high demand levels. For intersections with major and minor street, coordination conducted for major street had little impacts on the vehicles at the minor street.