Details
| Original language | English |
|---|---|
| Title of host publication | 2019 IEEE Intelligent Transportation Systems Conference (ITSC) |
| Pages | 3898-3905 |
| Number of pages | 8 |
| ISBN (electronic) | 978-1-5386-7024-8 |
| Publication status | Published - 2019 |
| Event | 2019 IEEE Intelligent Transportation Systems Conference, ITSC 2019 - Auckland, New Zealand Duration: 27 Oct 2019 → 30 Oct 2019 |
Abstract
Deep neural networks can be powerful tools, but require careful application-specific design to ensure that the most informative relationships in the data are learnable. In this paper, we apply deep neural networks to the nonlinear spatiotemporal physics problem of vehicle traffic dynamics. We consider problems of estimating macroscopic quantities (e.g., the queue at an intersection) at a lane level. First-principles modeling at the lane scale has been a challenge due to complexities in modeling social behaviors like lane changes, and those behaviors' resultant macro-scale effects. Following domain knowledge that upstream/downstream lanes and neighboring lanes affect each others' traffic flows in distinct ways, we apply a form of neural attention that allows the neural network layers to aggregate information from different lanes in different manners. Using a microscopic traffic simulator as a testbed, we obtain results showing that an attentional neural network model can use information from nearby lanes to improve predictions, and, that explicitly encoding the lane-to-lane relationship types significantly improves performance. We also demonstrate the transfer of our learned neural network to a more complex road network, discuss how its performance degradation may be attributable to new traffic behaviors induced by increased topological complexity, and motivate learning dynamics models from many road network topologies.
ASJC Scopus subject areas
- Computer Science(all)
- Artificial Intelligence
- Decision Sciences(all)
- Management Science and Operations Research
- Physics and Astronomy(all)
- Instrumentation
- Social Sciences(all)
- Transportation
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
2019 IEEE Intelligent Transportation Systems Conference (ITSC). 2019. p. 3898-3905.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research
}
TY - GEN
T1 - Neural-attention-based deep learning architectures for modeling traffic dynamics on lane graphs
AU - Wright, Matthew M.
AU - Ehlers, Simon Friedrich Gerhard
AU - Horowitz, Roberto
N1 - Funding Information: This research was supported by the National Science Foundation under grant CNS-1545116 and by Berkeley Deep-Drive. We also made use of the Savio computational cluster provided by the Berkeley Research Computing program at the University of California, Berkeley.
PY - 2019
Y1 - 2019
N2 - Deep neural networks can be powerful tools, but require careful application-specific design to ensure that the most informative relationships in the data are learnable. In this paper, we apply deep neural networks to the nonlinear spatiotemporal physics problem of vehicle traffic dynamics. We consider problems of estimating macroscopic quantities (e.g., the queue at an intersection) at a lane level. First-principles modeling at the lane scale has been a challenge due to complexities in modeling social behaviors like lane changes, and those behaviors' resultant macro-scale effects. Following domain knowledge that upstream/downstream lanes and neighboring lanes affect each others' traffic flows in distinct ways, we apply a form of neural attention that allows the neural network layers to aggregate information from different lanes in different manners. Using a microscopic traffic simulator as a testbed, we obtain results showing that an attentional neural network model can use information from nearby lanes to improve predictions, and, that explicitly encoding the lane-to-lane relationship types significantly improves performance. We also demonstrate the transfer of our learned neural network to a more complex road network, discuss how its performance degradation may be attributable to new traffic behaviors induced by increased topological complexity, and motivate learning dynamics models from many road network topologies.
AB - Deep neural networks can be powerful tools, but require careful application-specific design to ensure that the most informative relationships in the data are learnable. In this paper, we apply deep neural networks to the nonlinear spatiotemporal physics problem of vehicle traffic dynamics. We consider problems of estimating macroscopic quantities (e.g., the queue at an intersection) at a lane level. First-principles modeling at the lane scale has been a challenge due to complexities in modeling social behaviors like lane changes, and those behaviors' resultant macro-scale effects. Following domain knowledge that upstream/downstream lanes and neighboring lanes affect each others' traffic flows in distinct ways, we apply a form of neural attention that allows the neural network layers to aggregate information from different lanes in different manners. Using a microscopic traffic simulator as a testbed, we obtain results showing that an attentional neural network model can use information from nearby lanes to improve predictions, and, that explicitly encoding the lane-to-lane relationship types significantly improves performance. We also demonstrate the transfer of our learned neural network to a more complex road network, discuss how its performance degradation may be attributable to new traffic behaviors induced by increased topological complexity, and motivate learning dynamics models from many road network topologies.
UR - http://www.scopus.com/inward/record.url?scp=85076822708&partnerID=8YFLogxK
U2 - 10.48550/arXiv.1904.08831
DO - 10.48550/arXiv.1904.08831
M3 - Conference contribution
SN - 978-1-5386-7025-5
SP - 3898
EP - 3905
BT - 2019 IEEE Intelligent Transportation Systems Conference (ITSC)
T2 - 2019 IEEE Intelligent Transportation Systems Conference, ITSC 2019
Y2 - 27 October 2019 through 30 October 2019
ER -