基于车辆身份检测数据的车辆出行排放知识图谱研究

为实现个体车辆出行、排放行为的精细表征与挖掘，基于宣城市中心城区全量个体车辆的出行轨迹、技术参数、排放轨迹等多维交通大数据，以表征个体车辆出行过程的排放信息为主线，设计并构建车辆出行排放知识图谱.研究表明：①知识图谱直观地表征了“车辆-道路-出行-排放”信息的时空关联，可实现个体车辆在不同日期、不同时段、不同路段等多尺度出行特征的精细挖掘.以某辆小型客车为例，检索发现周一、周三出行主要连接的小时实体都为7：00、8：00和17：00，周五、非工作日出行连接的小时实体具有明显的随机性；周一、周三出行所连接的道路实体较少且基本一致，在宣水路、昭亭北路、昭亭南路的出行里程之和占比为63%~68%，周五、非工作日出行连接的道路实体则较为分散.②通过出行信息类、排放信息类实体的关联检索，可实现个体车辆出行排放时空特征的精细辨识和溯源分析.示例车辆的检索结果表明：周一车辆的CO日排放量为1.2g，是周六的2.5倍，同时在早高峰时段（7：00），车辆出行在交通繁忙路段时，伴随低水平车速，排放强度相对较高.

Research on vehicle travel emission knowledge graph based on vehicle identification data

To realize the refined representation and mining of individual vehicle travel and emission behaviour, a vehicle travel emission knowledge graph was established based on multi-dimensional traffic big data such as trajectories, technical parameters, and emission trajectories of all individual vehicles in the central urban area of Xuancheng City. Spatiotemporal correlations among vehicle, road, travel and emission information were intuitively represented by knowledge graph, achieving the fine mining of multi-scale travel characteristics of individual vehicles on different days, different time periods, and different road sections. A private passenger car was cited as an example, its hourly nodes connected to travel nodes on Monday and Wednesday were 7:00, 8:00 and 17:00, and hourly nodes connected to travel on Friday and non-working days were obviously random. The number of road nodes connected to travel on Monday and Wednesday were few and basically same, and the proportions of mileage on Xuanshui Road, Zhaoting North Road and Zhaoting South Road were 63%~68%, while the road nodes connected to travel on Fridays and non-working days were more dispersed. Through the associated retrieval of travel information and emission information nodes, fine identification and traceability analysis of the spatiotemporal characteristics of individual vehicle travel emissions could be realized. The retrieval results of the example vehicle showed that:the daily CO emission of the vehicle on Monday was 1.2g, which was 2.5times that of Saturday. During the morning peak hour (7:00), when the vehicle travelled on the busy road section, with the low level of vehicle speed, its emission intensity was relatively high.