Avoiding traffic congestion phenomena is an important aspect of efficient transportation infrastructure (e.g. toll roads) management. Traffic congestion phenomena can be avoided by forecasting volume traffic data. This paper aims to analyze how specific factors and parameters affect the behavior of traffic volume, like vehicle category, date, and weather data at a given timestamp and place. Moreover, the procedure of data prepossessing is presented to produce a cleaner data set that gives more fundamental information. Subsequently, spatio-temporal toll road prediction is achieved through a multi-layer perceptron. Finally, the proposed low-cost method is evaluated using real-life data from a toll plaza, while the experimental results show the efficiency of the proposed method.
MLP for Spatio-Temporal Traffic Volume Forecasting
2021-04-21
692226 byte
Conference paper
Electronic Resource
English
Spatio‐temporal adaptive graph convolutional networks for traffic flow forecasting
Wiley | 2023
|Spatio‐temporal adaptive graph convolutional networks for traffic flow forecasting
DOAJ | 2023
|Spatio-Temporal Graph Attention Convolution Network for Traffic Flow Forecasting
Transportation Research Record | 2024
|