【摘要】 为解决车辆在信号交叉口频繁停走导致高能耗的问题，同时考虑到自动驾驶车(autonomous vehicle, AV)与人工驾驶车(human-driven vehicle, HDV)混合交通的趋势，采用通过自动驾驶车引领人工驾驶车组成混合车队的方式实现信号交叉口处的生态驾驶。在混合车队建模方面，不仅考虑跟驰行为和能量消耗模型，还通过真实车辆轨迹数据进行理论分析，并采用Lighthill, Whitham and Richards (LWR)模型研究排队消散行为，为后续生态驾驶策略设计提供理论基础；研究了前方不同排队情况下目标速度的计算方法，提出了两阶段速度策略，并针对混合车队在同一绿灯时长内不能通过交叉口的情况设计了车队拆分策略。研究结果表明：在前方无排队的情况下，混合车队总能量消耗较自由驾驶模型减少了33.96%,比传统生态驾驶模型多节约了3.33%;在前方有排队且可消散的情况下，未考虑拆分策略的混合车队总能量消耗较自由驾驶模型减少了23.26%,应用拆分策略后总能量消耗的节约比例提高了4.41%;在存在二次排队的情况下，提出的模型相较自由驾驶模型总能量消耗下降了14.55%。研究结果有助于交通管理者根据不同交通状态为混合车队设计更加精准、灵活的动态策略，以降低单位车辆能耗，为实现双碳目标奠定基础。更多还原

【Abstract】 To address the high energy consumption resulting from frequent stopping and starting at signalized intersections, and in light of the increasing prevalence of mixed traffic involving autonomous vehicles(AVs) and human-driven vehicles(HDVs), a strategy was developed to facilitate eco-driving at intersections. This strategy was employed to utilize AVs to lead HDVs, thus forming mixed platoons. In modeling mixed vehicle platoons, both the following behavior and energy consumption models were considered, complemented by theoretical analyses based on real vehicle trajectory data. The LWR model was employed to investigate the queue dissipation behaviors, laying the theoretical groundwork for the design of subsequent eco-driving strategies. The method for calculating target speeds across various queuing scenarios were examined, and a two-stage speed strategy was proposed. Furthermore, a platoon-splitting strategy was developed for instances where mixed platoons cannot clear the intersection within a single green light phase. The results show that in scenarios without queuing, the total energy consumption of the mixed vehicle platoon reduced by 33.96% compared to the free-driving model, yielding an additional 3.33% savings over the traditional eco-driving model. In cases involving dissipatable queuing, the total energy consumption of the mixed platoon, excluding the splitting strategy, decreases by 23.26% relative to the free-driving model. The application of the splitting strategy further enhances energy savings by an additional 4.41%. In scenarios featuring secondary queuing, the proposed model achieves a 14.55% reduction in total energy consumption compared to the free-driving model. These findings provide valuable insights for traffic managers to devise more precise and flexible dynamic strategies for mixed platoons under varying traffic conditions, thereby reducing the per-vehicle energy consumption and supporting dual carbon objective. 8 tabs, 8 figs, 26 refs.更多还原