One Approach to Quantifying Rainfall Impact on the Traffic Flow of a Specific Freeway Segment

Abstract

Spatial constraints in urban areas very often lead to the application of traffic management measures to meet transport demands. Accordingly, it is very important to identify all potential impacts that could lead to reductions in the street network’s capacity. One such impact is weather conditions. The main motivation of this research is to analyze the impacts of rainfall on one of the most important segments of Belgrade’s street network that represents part of a freeway passing the city center. Our focus is on quantifying the impact of rainfall on speed and capacity. This paper proposes a new approach to analyzing the rainfall impact at the traffic lane level that provides additional possibilities to apply traffic management measures on the traffic lane level instead of at the direction level only. Functional dependences and differences in capacity and speed on traffic lanes were found under the influence of different rainfall intensities. Reductions in free-flow speed depend on the rain category and traffic lanes and vary from 4.5% to 11.58%, while reductions in traffic lane capacity range from 2.46% to 12.97%. We demonstrate the importance of considering the impact of rainfall at the traffic lane level, which could be quality input data for defining appropriate traffic management measures to mitigate the negative impacts.

Regional Expressway Freight Volume Prediction Algorithm Based on Meteorological Information

“…Qiao et al [38] proposed a short-term traffic volume prediction method based on a convolutional neural network and long short-term memory (DCNN-LSTM), which achieved good prediction results. Vidas et al [39] applied and compared four different machine learning algorithms (i.e. DT, RF, XGBoost and LSTM) for short-term travel time prediction.…”