事件状态下快速路行程时间预测研究

发布时间：2018-07-28 13:19

【摘要】：行程时间作为出行者最关注的交通参数之一,对于安排每日工作、生活计划具有重要意义。然而由于交通事件的不确定性、不可预知性,特别是在交通量日益增长的城市快速路上,轻微的交通事件就有可能引发大面积的行车延误,给居民生活工作带来较大不便。因此,通过研究事件状态下交通流特点、建立事件状态下快速路行程时间预测模型;构建行程时间预测、信息发布、交通诱导智能交通运输系统子系统对于降低出行延误具有重要意义。首先,文章从事件状态下快速路通行能力、交通流排队消散统计特性以及行程时间预测方法三个方面叙述了事件状态下快速路行程时间预测的基本理论与方法。其次,通过分析当前行程时间预测方法的优缺点,从事件状态下交通流统计特性出发,建立了波动理论-BP神经网络事件状态下快速路行程时间预测组合模型。最后,以北京市四环路交通事件历史统计资料为依据,根据路段上各环形线圈之间的行程时间计算路段实际行程时间;将事件发生时刻前后各36小时的交通量、地点车速、各事件阶段统计时间代入预测模型计算路段预测行程时间。利用绝对平均误差、均方根误差、平均绝对百分误差三个指标进行模型评价,结果表明所建立的组合预测模型较单一预测模型有较高的预测精度。将事件状态下行程时间预测模型应用在智能交通运输系统中,构建了智能交通运输系统事件状态下快速路行程时间预测子系统基本框架。
[Abstract]:As one of the most concerned traffic parameters, travel time is of great significance for daily work and life planning. However, due to the uncertainty and unpredictability of traffic events, especially on the urban expressway with increasing traffic volume, minor traffic events may lead to a large area of traffic delays and bring great inconvenience to the residents' life and work. Therefore, by studying the characteristics of traffic flow in the event state, establishing the expressway travel time prediction model under the event state, constructing the travel time prediction, releasing the information, The traffic guidance intelligent transportation system subsystem is very important to reduce travel delay. Firstly, this paper describes the basic theory and method of expressway travel time prediction in event state from three aspects: the capacity of expressway under event state, the statistical characteristics of traffic flow queuing dissipation and the method of travel time prediction. Secondly, by analyzing the advantages and disadvantages of the current travel time prediction methods, and starting from the statistical characteristics of traffic flow under the event state, a combined model for the prediction of the expressway travel time under the event state based on the fluctuation theory and BP neural network is established. Finally, based on the historical statistical data of traffic events on the fourth Ring Road in Beijing, the actual travel time of the road sections is calculated according to the travel time between the loops on the road sections. The traffic volume and the local speed of 36 hours before and after the incident are calculated. The statistical time of each event stage is substituted into the prediction model to calculate the predicted travel time of road sections. The model is evaluated by using three indexes: absolute mean error, root mean square error and average absolute percent error. The results show that the combined prediction model has higher prediction accuracy than the single prediction model. The travel time prediction model in the event state is applied to the intelligent transportation system, and the basic framework of the expressway travel time prediction subsystem in the event state of the intelligent transportation system is constructed.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491

【参考文献】