基于多源时空数据融合的车辆行为评价方法研究

Study on Evaluation Method of Vehicle Behavior Based on Multi-source Spatio-temporal Data Fusion

随着路网规模不断扩大,疲劳驾驶、超速、违停等行为频发,对车辆行为监管提出更高要求。基于车道收费数据、ETC门架计费数据、ETC门架牌识数据等,开展了多源数据融合研究,基于易发安全事故路段和车辆行驶轨迹、驾驶时长等数据进行了深度分析,提出了一种多参量动态信用评价方法,以有效评估车辆驾驶行为与交通安全事故发生的关联关系。

With the continuous expansion of the road network and the frequent occurrence of the fatigue driving, over-speed and illegal parking, there is a higher requirement for the supervision of vehicle behaviors.Based on lane toll data, ETC portal billing data, and ETC portal identification data, this paper carried out the in-depth research and analysis of the multi-source spatio-temporal data fusion according to the road sections with frequent traffic accidents and data of vehicle trajectory and driving time, and proposed a multi-parameter dynamic credit evaluation method to effectively evaluate the relationship between driving behaviors and traffic accidents.