Electric vehicle is a kind of new energy vehicle which uses batteries as energy supply unit.A huge gap in charging infrastructures will be created by the expansion of electric vehicles.The effectiveness and rationalit...Electric vehicle is a kind of new energy vehicle which uses batteries as energy supply unit.A huge gap in charging infrastructures will be created by the expansion of electric vehicles.The effectiveness and rationality of charging facilities will directly affect the convenience and economy of the users,as well as the safe operation of the power grid.Three types of charging facilities:charging pile,charging station and battery swap station are introduced in this paper.According to the different methods of charging infrastructure planning,the research status of the method of determining charging demand points is expounded.And the spatial distribution of charging demand points extracted by the current site selection method has a certain deviation.Then the models and algorithms of charging infrastructure optimized layout are reviewed.Currently,many researches focus on three categories optimization objectives:benefit of power company side,investment cost of charging facility and user side cost,and the genetic algorithm and particle swarm optimization are the main solving algorithms.Finally,the relative methods and development trend of the charging infrastructures optimized layout are summarized,and some suggestions on the optimized layout of electric vehicle charging infrastructures are given forward.展开更多
针对在信号交叉口前由于车辆转向和换道操作频繁容易引发冲突、降低通行效率的问题,构建基于深度强化学习(DQN)的车辆群体控制模型,优化车辆车道选择.首先,利用传感器和网联设备等获取周围车辆及交叉口信号灯实时状态信息,基于共享DQN...针对在信号交叉口前由于车辆转向和换道操作频繁容易引发冲突、降低通行效率的问题,构建基于深度强化学习(DQN)的车辆群体控制模型,优化车辆车道选择.首先,利用传感器和网联设备等获取周围车辆及交叉口信号灯实时状态信息,基于共享DQN模型进行车道选择,并根据该结果计算下一时刻位置、速度和转向角;进一步以效率及安全性指标建立奖励函数对车道选择决策实施评价,将状态信息、决策信息及奖励评价信息整合形成经验,存入同一经验池用于共享DQN模型参数迭代更新;最后,使用SUMO(simulation of urban mobility)与Python联合仿真搭建不同交通流量环境对训练后的模型进行验证.研究表明:相较于SUMO中的车道选择模型,基于共享DQN模型的信号交叉口前车辆群体车道选择模型,在低、中、高流量测试场景的平均速度均有提高,交叉口前排队长度分别减少了9.6%、22.5%和24.8%.本文模型可以有效减少信号交叉口的排队长度、提高信号交叉口前的路段平均速度、增强车辆从上游到达交叉口的效率,为未来车路协同的应用提供理论借鉴和技术支持.展开更多
交通基础设施对城市经济发展具有强大的推动作用,尤其是大型交通基础设施的建设,对土地利用和空间形态产生重大影响。针对城市交通基础设施广义经济效益评估分析需求,本文提出基于城市经济-土地利用-交通整体规划(Urban-Integrated Econ...交通基础设施对城市经济发展具有强大的推动作用,尤其是大型交通基础设施的建设,对土地利用和空间形态产生重大影响。针对城市交通基础设施广义经济效益评估分析需求,本文提出基于城市经济-土地利用-交通整体规划(Urban-Integrated Economy, Land Use, and Transport, U-IELUT)建模的交通基础设施广义经济效益评估框架和方法。在传统“四阶段”交通规划模型的基础上,加入城市经济和人口预测、社会经济活动空间分配、城市空间开发,以及交通基础设施广义经济效益评估模块,构建面向交通基础设施广义经济效益评估的“PECAS+(Production, Exchange and Consumption Allocation System)”广义经济效益评估模型。以武汉地铁2号线为例,基于武汉“PECAS+”广义经济效益评估模型测算其直接经济效益和广义经济效益。结果表明,地铁2号线在2027年的直接经济效益约为10.43亿元;广义经济效益中,动态集聚效益约为2.64亿元,约为直接经济效益的25.3%,说明广义经济效益尤其是集聚效益等带来的经济效益不可忽略。同时,还探明交通基础设施建设对不同小区的影响差异,即广义经济效益的空间分布,可为交通基础设施投资建设提供多维度决策支持。展开更多
基金Project(21805217)supported by the National Natural Science Foundation of ChinaProject(2015BAG08B02)supported by the National Key Technologies Research and Development Program of ChinaProject(2019IVB014)supported by the Fundamental Research Funds for the Central Universities,China。
文摘Electric vehicle is a kind of new energy vehicle which uses batteries as energy supply unit.A huge gap in charging infrastructures will be created by the expansion of electric vehicles.The effectiveness and rationality of charging facilities will directly affect the convenience and economy of the users,as well as the safe operation of the power grid.Three types of charging facilities:charging pile,charging station and battery swap station are introduced in this paper.According to the different methods of charging infrastructure planning,the research status of the method of determining charging demand points is expounded.And the spatial distribution of charging demand points extracted by the current site selection method has a certain deviation.Then the models and algorithms of charging infrastructure optimized layout are reviewed.Currently,many researches focus on three categories optimization objectives:benefit of power company side,investment cost of charging facility and user side cost,and the genetic algorithm and particle swarm optimization are the main solving algorithms.Finally,the relative methods and development trend of the charging infrastructures optimized layout are summarized,and some suggestions on the optimized layout of electric vehicle charging infrastructures are given forward.
文摘针对在信号交叉口前由于车辆转向和换道操作频繁容易引发冲突、降低通行效率的问题,构建基于深度强化学习(DQN)的车辆群体控制模型,优化车辆车道选择.首先,利用传感器和网联设备等获取周围车辆及交叉口信号灯实时状态信息,基于共享DQN模型进行车道选择,并根据该结果计算下一时刻位置、速度和转向角;进一步以效率及安全性指标建立奖励函数对车道选择决策实施评价,将状态信息、决策信息及奖励评价信息整合形成经验,存入同一经验池用于共享DQN模型参数迭代更新;最后,使用SUMO(simulation of urban mobility)与Python联合仿真搭建不同交通流量环境对训练后的模型进行验证.研究表明:相较于SUMO中的车道选择模型,基于共享DQN模型的信号交叉口前车辆群体车道选择模型,在低、中、高流量测试场景的平均速度均有提高,交叉口前排队长度分别减少了9.6%、22.5%和24.8%.本文模型可以有效减少信号交叉口的排队长度、提高信号交叉口前的路段平均速度、增强车辆从上游到达交叉口的效率,为未来车路协同的应用提供理论借鉴和技术支持.
文摘交通基础设施对城市经济发展具有强大的推动作用,尤其是大型交通基础设施的建设,对土地利用和空间形态产生重大影响。针对城市交通基础设施广义经济效益评估分析需求,本文提出基于城市经济-土地利用-交通整体规划(Urban-Integrated Economy, Land Use, and Transport, U-IELUT)建模的交通基础设施广义经济效益评估框架和方法。在传统“四阶段”交通规划模型的基础上,加入城市经济和人口预测、社会经济活动空间分配、城市空间开发,以及交通基础设施广义经济效益评估模块,构建面向交通基础设施广义经济效益评估的“PECAS+(Production, Exchange and Consumption Allocation System)”广义经济效益评估模型。以武汉地铁2号线为例,基于武汉“PECAS+”广义经济效益评估模型测算其直接经济效益和广义经济效益。结果表明,地铁2号线在2027年的直接经济效益约为10.43亿元;广义经济效益中,动态集聚效益约为2.64亿元,约为直接经济效益的25.3%,说明广义经济效益尤其是集聚效益等带来的经济效益不可忽略。同时,还探明交通基础设施建设对不同小区的影响差异,即广义经济效益的空间分布,可为交通基础设施投资建设提供多维度决策支持。
