Short-Term Traffic Flow Forecasting for Freeway Incident-Induced Delay Estimation

Freeway incidents not only threaten travelers’ safety but also cause severe congestion. Incident-induced delay (IID) refers to the extra travel delay resulting from incidents on top of the recurrent congestion. Quantifying IID would help people better understand the real cost of incidents, maximize the benefit-cost ratio of investment on incident remedy actions, and develop active traffic management and integrated corridor management strategies. By combining a modified queuing diagram and short-term traffic flow forecasting techniques, this study proposes an approach to estimate the temporal IID for a roadway section, given that the incidents occurs between two traffic flow detectors. The approach separates IID from the total travel delay, estimates IID for each individual incident, and only takes volume as input for IID quantification, avoiding using speed data that are widely involved in previous algorithms yet are less available or prone to poor data quality. Therefore, this approach can be easily deployed to broader ranges where only volume data are available. To verify its estimation accuracy, this study captures two incident videos and extracts ground-truth IID data, which is rarely done by previous studies. The verification shows that the IID estimation errors of the proposed approach are within 6% for both cases. The approach has been implemented in a Web-based system, which enables quick, convenient, and reliable freeway IID estimation in the Puget Sound region in the state of Washington.     

公路隧道交通量的预测及分析

公路隧道交通量的预测对隧道通风系统的节能以及降低隧道运营成本有很重要的意义。分别利用多元统计分析法和BP神经网络两种方法对公路隧道交通量进行了预测,并对两类预测数据进行分析比较,得出了多元统计分析法适用于车流量少而且稳定的公路隧道的预测,而BP神经网络法则适用于车流量大而且不稳定的公路隧道的预测的结论。     

交通流预测方法综述

实时交通流预测是智能运输系统研究的一个重要问题。为此,建立了许多预测模型,有历史平均模型、时间序列模型、卡尔曼滤波模型、非参数回归模型、神经网络模型和组合模型等。总结评述现存的各类模型,提出交通流预测研究领域今后可能的发展趋势。     

交通流缺失数据处理方法比较分析

针对交通数据的缺失问题,采用基于时间相关性、空间相关性和时空相关性的多种数据修复方法对缺失数据进行处理.基于时间相关性的修复方法包括历史数据法、移动平均法、指数平滑法和线性回归法等.基于空间相关性的修复方法利用相邻车道和相邻检测器所采集的数据对缺失值进行处理.基于时空相关性的数据修复方法结合交通流的时间相关性与空间相关性对缺失数据进行修复.基于美国加州I-880高速公路交通流数据的实验结果表明,平滑系数α=0.1时的指数平滑法和利用相邻车道数据加权平均法得到的缺失值修复结果最优.     

Data Fusion-Based Traffic Density Estimation and Prediction

Traffic congestion has become a major challenge in recent years in many countries of the world. One way to alleviate congestion is to manage the traffic efficiently by applying intelligent transportation systems (ITS). One set of ITS technologies helps in diverting vehicles from congested parts of the network to alternate routes having less congestion. Congestion is often measured by traffic density, which is the number of vehicles per unit stretch of the roadway. Density, being a spatial characteristic, is difficult to measure in the field. Also, the general approach of estimating density from location-based measures may not capture the spatial variation in density. To capture the spatial variation better, density can be estimated using both location-based and spatial data sources using a data fusion approach. The present study uses a Kalman filter to fuse spatial and location-based data for the estimation of traffic density. Subsequently, the estimated data are utilized for predicting density to future time intervals using a time-series regression model. The models were estimated and validated using both field and simulated data. Both estimation and prediction models performed well, despite the challenges arising from heterogeneous traffic flow conditions prevalent in India.     

基于混沌理论的高速公路网短时交通流量预测研究

随着高速公路网的建成及其交通流量的不断增大,对高速路网交通流实时控制和诱导服务的需求日益突出,而高速公路网短时交通流量的预测,不仅是交通流实时控制和诱导服务的基础和依据,而且预测结果的准确性对改善高速公路网的通行能力和服务水平有重要影响。基于混沌时间序列分析和预测的理论,建立了高速公路网短时交通流预测模型,计算给定区域高速公路网多断面短时交通流量预测值,结果表明利用多维混沌时间序列法预测高速公路网短时交通流量可行且具有较高的精度。     

短时交通流量两种预测方法的研究

实时、准确的完成短时交通流量预测是实现交通控制与诱导的关键。采用基于L-M算法的BP神经网络预测方法和基于混沌时间序列的预测方法对短时交通流量时间序列进行了预测研究,给出两种方法的基本原理及具体的预测步骤,并对一组实际的流量数据进行了预测。仿真结果表明:两种方法都能较准确的预测交通流量,但混沌时间序列方法的实时性更好一些,更适合于预测短时交通流量。     

基于路网空间拓扑结构的交通流数据填补模型研究

由于传感器故障、状况不稳定或数据传输等问题,交通流数据存在大量丢失的情况,对依靠这些数据进行交通科学决策产生了很大的影响.基于以上问题,提出一种基于空间拓扑结构的交通流数据填补方法,模型通过选取区域的典型道路训练上下游之间相关参数,根据道路类型匹配典型道路,应用典型道路的参数进行交通流数据填补,并利用武汉市实际交通检测器数据进行模型验证.选取武汉市实际道路上2个典型地磁检测器进行数据填补误差分析,不同填补方式下的平均相对误差为52.88%和51.93%,相对于传统的交通流数据填补建模来说,基于空间拓扑结构的填补模型在实际的应用过程中,能够在更大的范围内对丢失的交通流数据进行填补.     

基于ETC数据和A-BiLSTM神经网络的高速公路节假日短时交通流预测模型

电子不停车收费（electronic toll collection,ETC）门架系统为节假日高速公路短时交通流预测提供了数据支撑。针对节假日场景下高速公路交通流的非线性和复杂性特征,基于ETC门架数据研究了由注意力机制（attention）和双向长短期记忆（bidirectional long/short-term memory,BiLSTM）神经网络组成的Attention-BiLSTM（A-BiLSTM）组合模型。通过对ETC门架数据进行预处理,保证模型输入的可靠性;采用滑动窗口方法构建监督学习样本,提高模型学习效率。在模型中,使用BiLSTM神经网络,实现对交通流数据前向和后向时间依赖性特征的深入提取;引入注意力机制动态地权衡网络提取信息的重要程度,增强隐藏层特征的非线性表达能力;利用贝叶斯优化方法对模型进行超参数调优,提高模型的预测性能。采集大理-丽江高速公路白汉场至拉市镇的门架数据,处理成时间粒度为5,10,15 min的交通流数据进行模型验证。实验结果表明：①相比于自回归移动平均模型、支持向量机的预测结果,A-BiLSTM组合模型的均方根误差（root mean square error,RMSE）分别降低了73.3%和49.1%,平均绝对误差（mean absolute error,MAE）分别降低了76.0%和56.3%,预测效果好,可应用于实际的交通运营管理。②相比于未引入注意力机制的BiLSTM,A-BiLSTM组合模型的RMSE降低了41.9%,MAE降低了46.0%。③A-BiLSTM组合模型在5 min的时间粒度下表现最好,与输入数据时间粒度为10,15 min情况下所构建的模型预测误差相比,RMSE分别降低34.5%和42.1%,MAE分别降低39.9%和46.3%。