Learning Traffic Speed Dynamics from Visualizations
arXiv:2105.01423 · doi:10.1109/ITSC48978.2021.9564541
Abstract
Space-time visualizations of macroscopic or microscopic traffic variables is a qualitative tool used by traffic engineers to understand and analyze different aspects of road traffic dynamics. We present a deep learning method to learn the macroscopic traffic speed dynamics from these space-time visualizations, and demonstrate its application in the framework of traffic state estimation. Compared to existing estimation approaches, our approach allows a finer estimation resolution, eliminates the dependence on the initial conditions, and is agnostic to external factors such as traffic demand, road inhomogeneities and driving behaviors. Our model respects causality in traffic dynamics, which improves the robustness of estimation. We present the high-resolution traffic speed fields estimated for several freeway sections using the data obtained from the Next Generation Simulation Program (NGSIM) and German Highway (HighD) datasets. We further demonstrate the quality and utility of the estimation by inferring vehicle trajectories from the estimated speed fields, and discuss the benefits of deep neural network models in approximating the traffic dynamics.
arXiv comment8 pages, 9 figures; Submitted to the 2021 IEEE International Intelligent Transportation Systems Conference (ITSC)
References in corpus (9)
- Deep Learning for Short-Term Traffic Flow Prediction
- Reconstructing the Traffic State by Fusion of Heterogeneous Data
- Position weighted backpressure intersection control for urban networks
- Multi-scale Perimeter Control Approach in a Connected-Vehicle Environment
- Traffic Data Imputation using Deep Convolutional Neural Networks
- Sparse Travel Time Estimation from Streaming Data
- Learning Traffic Flow Dynamics using Random Fields
- Learning-based Traffic State Reconstruction using Probe Vehicles
- Nonlinear Traffic Prediction as a Matrix Completion Problem with Ensemble Learning