随着车联网和无人驾驶技术的不断深入发展和广泛使用，城市交通系统必将转变为由自动网联车与人工驾驶网联车组成的异构交通系统。在上述背景下，本项目提出将可调控的自动网联车作为新的车流调控手段，研究其驾驶行为对路段随机车流的影响，旨在提高异构交通流调控能力和改善路口交通性能。本项目从异构车流特征出发，通过数据和机理分析不同数量和车速变化下自动网联车对路段局部车流的影响；考虑不同类型车辆在交通需求量和分布比例上的随机性，建立包含自动网联车车速影响的随机异构交通流模型；面对路段自动网联车位置和数量的不确定性，提出概率模型在线选择受控自动网联车构型，以多车协同群控方式对路段车速进行精准调控；结合路段车速控制和信号灯控制，构建异构交通环境下路口高效车灯协同随机优化方法。本项目研究抓住当前社会和科技发展所带来的新问题和新挑战，着眼于解决异构交通环境下车流调控问题，成果具有较强的学术意义和应用价值。

With the continuous development and wide use of the Internet of vehicles and driverless vehicle technology, the urban transportation system will be transformed into a heterogeneous transportation system composed of connected autonomous vehicles and connected human-driven vehicles. In this context, our project takes the adjustable connected autonomous vehicle as a new means for the traffic flow control and studies the influence of its velocity change on the link traffic flow, so as to enhance the regulation ability of heterogeneous traffic flow and improve the traffic performance at intersections. The influence of connected autonomous vehicles with different numbers and velocity changes on the link traffic flow speed is first studied through the data and mechanism analysis in this project. Then a stochastic cell transmission model of heterogeneous traffic flow is established, which takes the velocity change of connected autonomous vehicles into consideration. Considering the uncertainty in the position and number of connected autonomous vehicles in a link, this project proposes a random probability model to select the appropriate controlled connected autonomous vehicles online and a multi-vehicle cooperative group control method to accurately regulate the link traffic flow speed. Combining with the speed control of link traffic flow and the traditional traffic light control, an efficient stochastic collaborative optimization method for the intersection control under heterogeneous traffic flow is constructed. This project focuses on the new problems and challenges brought by the current social and technological development, and aims at solving the difficult problem of traffic flow regulation in the heterogeneous traffic environment, which have strong academic significance and application value.