基于强化学习的干线信号混合协同优化方法

交通拥堵已成为很多大中城市普遍存在的社会问题。信号控制作为缓堵保畅的重要措施之一，愈发受到社会关注。信号优化手段可分为模型驱动和数据驱动两类，且随着交通大数据的不断充实，基于强化学习的数据驱动方法日益成为新兴发展方向。然而，现有数据驱动类研究主要偏重于决策模型设计，缺乏对智能体结构的探讨；同时，在多路口协同方面多采用分布式策略， 忽略了智能体之间信息交互，无法保障区域层面的整体最优性。为此，本文以干线信号为对象， 构建一种多智能体混合式协同决策的信号优化方法。首先，针对交通状态的多样性、异构性及数据不均衡性，设计分布训练-分区记忆的单智能体决策模型，并优化状态空间和回报函数，界定单路口控制的最佳方案；其次，融合分布式和集中式学习的模型优势设计多智能体交互方法，在单路口分布式控制的基础上，设置中心智能体评价局部智能体的决策行为并反馈附加回报以调整局部智能体的决策模型，实现干线多信号的协同运行。最后，搭建仿真平台完成效果测试与算法对比。结果表明：新方法与独立优化和分布式协同相比，在支路交通流基本不受影响的前提下， 干线停车次数分别降低了14.8%和13.6%，具有更好的控制效果。     

快速路与邻接交叉口分散换道和速度引导方法

研究车路协同城市快速路与邻接交叉口主线分散换道和速度引导自适应控制方法. 对高饱和度入口匝道与邻接交叉口，提出主线分散换道自适应控制方法，依据合流区上游不同车道密度制定换道规则，以主线流量最大化为目标确定邻接交叉口相位相序；对出口匝道存在超长排队，提出主线速度引导自适应控制方法，依据主线上游车辆目的地确定速度引导策略，以出口匝道需求与通行能力相匹配为目标确定出口匝道关联相位优先权. 采用元胞自动机模型仿真验证，结果表明，所提方法与非协调控制、传统协调控制、车路协同交叉口自适应控制相比，区域流量分别提高17.38%、5.52%、10.06%，总时间消耗分别下降35.86%、 26.21%、17.39%.     

Determinants of airport retail revenue: a review of literature

ABSTRACT

As the airport retail industry continues to grow due to increasing travel demands, airport operators are increasingly developing their retail revenue potential to ensure financial viability. This study aims to provide a review of airport retail literature and identify the salient factors associated with passenger shopping behaviour. The paper presents a review of contemporary airport retail literature, covering a total of 50 studies from 1998 to 2018. The review identified 26 factors, which could broadly be grouped into five categories: airport/operator related; passengers' demographic related; passengers’ travel related; passengers’ psychological related and passengers’ resources related. In addition to providing a summary of the statistically significant factors across studies, the review identifies and discusses potential approaches for future research. These include the importance of considering both “to spend or not” decision and how much to spend and how most empirical studies focussed on the former; the gap in empirical analysis on the impact of airport terminal design layout on retail performance; and the gap in application of general consumer shopping behavioural models to airport retail problems. The paper concludes with the suggestion that we can build on the existing studies to develop a hybrid approach to solve several of the identified gaps simultaneously.     

一种基于地图辅助的自动驾驶视-惯融合定位方法

视觉同步定位与建图(Simultaneous Localization and Mapping, SLAM)方法广泛应用于自动驾驶领域。传统的方法利用车载摄像头表征车辆周围环境，同时估计自身位置，当车辆运动过快时，定位精度和鲁棒性会下降。针对此问题，本文提出一种地图辅助的视-惯融合定位方法。该方法在ORB-SLAM2(Oriented FAST and Rotated BRIEF SLAM2)的基础上拓展地图保存功能，将建图和定位拆分为两个独立模块，车辆首先以较慢的速度构建并保存具有视觉特征的地图，然后，在第2次运行时车载计算机调用预先保存的地图实现精确且稳定的定位性能。由于构建地图阶段采用了图优化算法融合惯性测量单元(Inertial Measurement Unit, IMU)的信息，地图误差得到有效校正。在KITTI数据集场景和实际场景中验证了所提方法的良好性能。实验结果表明，所提方法在4, 8, 16 m·s-1 驾驶速度下的定位精度分别为2.59，2.61，2.73 m，图像失帧率和路径丢失率分别为3.76%和1.38%，3.89%和1.69%，4.27%和1.84%。相比原始的ORB-SLAM2方法，系统定位精度和鲁棒性均得到了提高。     

某平台机舱进排风口气流组织的CFD模拟

某自升式平台的机舱进、排风口位置比较接近,机舱排出的热气受大气风影响可能返回进风口,进而提高进风温度,成为一个设计隐患。文章应用CFD技术,通过模拟一定的大气环境条件下进排风口处的气流组织,找出最不利工况,并提出切实可行的改进建议。     

Economic factors affecting the registration of lower CO2 emitting aircraft in Europe

We examine the various forces influencing the development and uptake of environmentally beneficial technical changes, focusing on airline technology. Within this context, we consider not only the nature of competition within the final market in which aircraft, an intermediate product, are sold, but also that of the product market itself, the commercial airline industry. The reasons for the gradual reduction in CO₂ per seat per aircraft movement in aircraft design are examined in terms of the real costs of aviation fuel, changes in the nature of the supply industry, the movement towards carbon cap-trade policies, and endogenous technical progress in the technology of the industry. The latter being taken as an empirical proxy for the role market forms play in influencing the fuel efficiency of the types of aircraft used. The results support the existence of these latter forces on the demand for aircraft types, allowing for other influences that affect aircraft technology.     

客运枢纽"人群迷路区"的快速识别与优化 ——以北京南站为例

为缓解大型客运枢纽因诱导服务不到位而产生的人流拥堵问题，采用室内WiFi 定位技术和Mysql获取行人交通流分布热力图，通过百度地图坐标拾取系统识别人流聚集区的具体位置。通过定义迷路人群的交通特征参数，提出“人群迷路区”概念；应用视频识别追踪技术，分析行人速度与加速度的方差，于人流密集区运用视频识别技术寻找“人群迷路区”。最后以北京南站为例进行实地调研，利用调研数据识别北京南站的“人群迷路区”。基于人群速度与加速度方差，“人群迷路区”被分为四类：单纯人流聚集区、因标识不足而产生的人群迷路区、因标识过于复杂而产生的人群迷路区、因商业设施或障碍物遮挡而产生的人群迷路区。同时提出加大商业活动监管力度、简化部分引导标识、增加其他交通方式换乘区域标识等优化方法，可使人群行进效率提高约25%。     

有轨电车氢燃料电池系统通风散热方法研究

为解决有轨电车氢燃料电池系统的散热问题,文章提出一种新型车载氢燃料电池系统通风散热方法,分析了该方法的散热原理和控制策略,并对通风散热结构进行了详细介绍。该车载氢燃料电池系统通风散热方法提高了系统散热效率,降低了散热风机能耗。     

考虑对地航速和航向的船舶典型轨迹提取方法

基于船舶自动识别系统(Automatic Identification System, AIS)数据的船舶典型轨迹挖掘需要经过两个重要步骤，一是压缩 AIS 数据，二是聚类压缩后的 AIS 数据。传统的DP(DouglasPeucker)压缩算法，只考虑船舶轨迹的压缩形状，忽视了船舶航行中其他重要信息。为解决此问题，把对地航速和航向加入到DP算法的压缩过程中。在AIS轨迹聚类方面，传统谱聚类方法只对船舶轨迹的位置进行相似性度量，没有考虑船舶轨迹的其他维度，针对此问题，提出多属性轨迹相似性度量方法。由于不同的输入参数影响着最终的聚类质量，引入Calinski-Harabasz指标评价谱聚类算法，实现聚类参数的自适应选择。利用山东威海水域的实际AIS数据进行实例研究，并与传统谱聚类算法做比较实验。实验结果表明，利用该方法提取到的典型轨迹符合真实水域的交通情况，相较于传统谱聚类方法具有更高的聚类质量。     

Factors affecting public transportation usage rate: Geographically weighted regression

As the number of private vehicles grows worldwide, so does air pollution and traffic congestion, which typically constrain economic development. To achieve transportation sustainability and continued economic development, the dependency on private vehicles must be decreased by increasing public transportation usage. However, without knowing the key factors that affect public transportation usage, developing strategies that effectively improve public transportation usage is impossible. Therefore, this study respectively applies global and local regression models to identify the key factors of usage rates for 348 regions (township or districts) in Taiwan. The global regression model, the Tobit regression model (TRM), is used to estimate one set of parameters that are associated with explanatory variables and explain regional differences in usage rates, while the local regression model, geographically weighted regression (GWR), estimates parameters differently depending on spatial correlations among neighbouring regions. By referencing related studies, 32 potential explanatory variables in four categories, social-economic, land use, public transportation, and private transportation, are chosen. Model performance is compared in terms of mean absolute percentage error (MAPE) and spatial autocorrelation coefficient (Moran’ I). Estimation results show that the GWR model has better prediction accuracy and better accommodation of spatial autocorrelation. Seven variables are significantly tested, and most have parameters that differ across regions in Taiwan. Based on these findings, strategies are proposed that improve public transportation usage.