Search

At present, the congestion of highways in China is severe. Traffic flow prediction plays an important role in the intelligent transportation system. If it can achieve high-precision prediction, it will be able to efficiently manage traffic and alleviate congestion. To solve this issue, this paper proposed a multi-channel traffic flow prediction method (MCST-Transformer)considering spatiotemporal correlation. Firstly, it used Transformer to extract the internal laws of different data, and then introduced a spatial correlation module to mine the association features of different data. Finally, it integrated global information through channel attention. Using the data of Guangdong province highway, the proposed method realized the traffic flow prediction of 92 toll stations within two hours with high precision. The results show that MCST-Transformer is superior to traditional machine learning methods and time series models based on the attention mechanism. Under the prediction horizon of 120 min, MAPE decreases by 5. 1 % compared with Bayesian regression. Compared with deep learning algorithms like Seq2Seq-Att and Seq2Seq, the overall MAPE of the proposed method can also be reduced by 0. 5%. It indicates that the multi-channel approach can distinguish the characteristics of different data, so as to acquire better performance. [ABSTRACT FROM AUTHOR]

With the development of economy and urbanization in China, the issue of traffic congestion is increasingly prominent. Highways as the lifelines between cities are particularly vulnerable during the holidays. In order to manage highways more efficiently and regulate traffic more timely, it is necessary to detect congestion events on highways to realize the supervision of highway traffic. Using recognition data of highway path recognition system, transactions data of toll station, and GPS trajectory data of coach buses, touring buses and dangerous goods transport vehicles, this paper proposes a hierarchical congestion interval detection method which adopts fuzzy comprehensive evaluation algorithm to identify the traffic status of road sections. Taking the Changsha-Zhuzhou-Xiangtan urban agglomeration region of Hunan province as a case study, this paper uses the proposed method to detect and analyze traffic status of the experimental area in two weeks in February, 2018. [ABSTRACT FROM AUTHOR]

At present,many cities in China are short of pipeline-flow observations,which leads to large challenges to construct urban hydrological models for detailed rainfall-runoff simulations. In this study,the core district of Yizhuang Economic Development Zone was selected as the study area. Based on the accurate digitalized land-surface data,field survey and empirical verification of urban storm process was carried out in terms of the maximum runoff depth at specific spots,and the SWMM model was calibrated and verified for the two typical storm events in July 21st,2012 and June 23rd,2011. Subsequently,the runoff response rate and traffic congestion risk of design rainstorms in different return periods were analyzed and evaluated in sub-catchments with various land-uses. The results show that as return period increases,street and transportation land-uses have a higher response rate to the increase of rainstorm than park and green land-uses do,and the former has a higher risk of waterlogging-triggered traffic congestion than the latter. This finding indirectly reflects the variability of runoff response rates under different land-surface conditions. This study is potential to provide decision-makers with valuable suggestions on effective local urban-flood control and early warning. [ABSTRACT FROM AUTHOR]