面向新型混合车流的协同式自适应巡航控制方法研究

As the roll-out of the connected automation, CACC, as one of the most promising connected automation technology, is close to being commercialized. Along with the commercialization of CACC comes the mixed traffic of conventional vehicles and CACC vehicles. Previous studies, including findings from this research team, have shown that the emerging mixed traffic will have distinct features from conventional traffic due to the much shorter following distance between CACC vehicles. The changes in characteristics will include but are not limited to: car-following behavior, lane change behavior, and traffic flow fundamental diagram. The impact of introducing the state-of-the-art CACC into transportation system is both positive and negative, because CACC not only reduces headway and increase throughput, but also cause impedance to lane change maneuver of human drivers. However, no study has conducted a systematic evaluation on CACC to weight its pros against its cons. In addition, no CACC has been developed to optimize its performance in the context of such emerging mixed traffic. Therefore, this research team proposes to further advance the Cooperative Adaptive Cruise Control (CACC) in the context of partially Connected Automated Vehicles (CAV) environment. The ultimate goal of this study is to develop control algorithms for clustering and platooning of CACC vehicles and dissolution of CACC platoon. In order to realize this, car-following model, lane change model of both CACC vehicle and regular human driver under the influence of CACC need to be constructed and calibrated. Traffic flow theory of partially CAV traffic shall be introduced. Findings of this proposed study will lead CACC technology closer to implementation and serve as a foundation for other CAV applications under partially CAV environment.

协同式自适应巡航控制是网联环境下可最先实现的重要功能，对节能减排、效率提升和改善安全价值巨大。现有协同式自适应巡航控制研究多聚焦于纯粹网联自动车辆环境，缺乏常规车辆混入后的分析和对策。而实际上协同式自适应巡航车辆、常规车辆所组成的多模式混合交通流已呈现并将长期持续存在。本项目聚焦快速路环境，面向常规和协同式自适应巡航车辆共同构成的新型多模式混合交通流，研究多模式混合交通流中常规和协同式自适应巡航车辆的交通行为，解析异种车辆相互影响机理，构建微观交通流跟驰模型、换道模型。考虑车辆运行多目标，设计多模式混合交通流的编队形成、保持优化控制方法；针对主动和被动的驶离车队、穿越车队等需求，建立混合车队拆分优化控制方法。设计通信系统、车辆和交通流环境在环仿真平台进行控制策略评价分析。研究成果将为多模式混合交通流的控制提供理论基础，为协同式自适应巡航的落地应用提供理论指导，具有重要的理论意义和实用价值。

网联自动驾驶技术的研发及其应用已成为国内外智能交通运输系统发展的新趋势。其中，协同式自适应巡航控制（CACC: Cooperative Adaptive Cruise Control）被认为是最容易落地应用的网联自动驾驶技术。CACC通过车车协同缩短跟驰间距，减小车辆速度波动，提升车队的稳定性，是一项重要的网联自动驾驶功能，既可以提升行驶的安全性，也能够改善交通效率，降低行驶能耗。本项目聚焦于协同式自适应巡航车辆和常规人工控制车辆混合交通流条件下的协同自适应巡航控制方法，通过解析多模式车辆间驾驶行为影响机理，深化对新型混合交通流驾驶行为的理解；通过建立多模式车辆混合交通流背景下的车流编队优化控制方法、编队拆分控制方法，提升了新型混合交通流管控能力，并促进混合交通流中网联自动驾驶技术的发展。.在完成项目任务书既定任务之外，本项目扩展研究了面向新型混合交通流的智能网联车辆决策控制方法、智能网联多车多目标耦合决策控制方法以及智能网联车辆仿真验证方法，从而有效保证并支撑了面向新型混合车流的协同式自适应巡航控制方法相关研究。.本项目在申请和实施过程中，始终注意跟踪国内外新理论、新方法和新技术的发展，将基础理论研究和实际应用相结合，因而取得了具有实用价值的理论成果。其中，共发表发明专利5项，发表文章20篇，极大丰富了协同式自适应巡航控制的基本理论和研究方法，具有重大科学意义和学术价值。