Reducing Urban Traffic Congestion Using Deep Learning and Model Predictive Control

Zhun Yin; Tong Liu; Chieh Wang; Hong Wang; Zhong Ping Jiang

doi:10.1109/TNNLS.2023.3264709

Reducing Urban Traffic Congestion Using Deep Learning and Model Predictive Control

Zhun Yin, Tong Liu, Chieh Wang, Hong Wang, Zhong Ping Jiang

Research output: Contribution to journal › Article › peer-review

9 Scopus citations

Abstract

This article proposes a deep learning (DL)-based control algorithm - DL velocity-based model predictive control (VMPC) - for reducing traffic congestion with slowly time-varying traffic signal controls. This control algorithm consists of system identification using DL and traffic signal control using VMPC. For the training process of DL, we established a modeling error entropy loss as the criteria inspired by the theory of stochastic distribution control (SDC) originated by the fourth author. Simulation results show that the proposed algorithm can reduce traffic congestion with a slowly varying traffic signal control input. Results of an ablation study demonstrate that this algorithm compares favorably to other model-based controllers in terms of prediction error, signal varying speed, and control effectiveness.

Original language	English
Pages (from-to)	12760-12771
Number of pages	12
Journal	IEEE Transactions on Neural Networks and Learning Systems
Volume	35
Issue number	9
DOIs	https://doi.org/10.1109/TNNLS.2023.3264709
State	Published - 2024

Funding

This work was supported in part by the U.S. Department of Energy, Vehicle Technologies Office, Energy Efficient Mobility Systems Program, and in part by the National Science Foundation under Grant EPCN-1903781.

Keywords

Deep learning (DL)
gain-scheduling
model predictive control (MPC)
traffic signal control
velocity-based linearization (VL)

Access to Document

10.1109/TNNLS.2023.3264709

Cite this

@article{a64e0e43ad2b4cbcb2911a2ec4181969,

title = "Reducing Urban Traffic Congestion Using Deep Learning and Model Predictive Control",

abstract = "This article proposes a deep learning (DL)-based control algorithm - DL velocity-based model predictive control (VMPC) - for reducing traffic congestion with slowly time-varying traffic signal controls. This control algorithm consists of system identification using DL and traffic signal control using VMPC. For the training process of DL, we established a modeling error entropy loss as the criteria inspired by the theory of stochastic distribution control (SDC) originated by the fourth author. Simulation results show that the proposed algorithm can reduce traffic congestion with a slowly varying traffic signal control input. Results of an ablation study demonstrate that this algorithm compares favorably to other model-based controllers in terms of prediction error, signal varying speed, and control effectiveness.",

keywords = "Deep learning (DL), gain-scheduling, model predictive control (MPC), traffic signal control, velocity-based linearization (VL)",

author = "Zhun Yin and Tong Liu and Chieh Wang and Hong Wang and Jiang, {Zhong Ping}",

note = "Publisher Copyright: {\textcopyright} 2012 IEEE.",

year = "2024",

doi = "10.1109/TNNLS.2023.3264709",

language = "English",

volume = "35",

pages = "12760--12771",

journal = "IEEE Transactions on Neural Networks and Learning Systems",

issn = "2162-237X",

publisher = "Institute of Electrical and Electronics Engineers",

number = "9",

}

TY - JOUR

T1 - Reducing Urban Traffic Congestion Using Deep Learning and Model Predictive Control

AU - Yin, Zhun

AU - Liu, Tong

AU - Wang, Chieh

AU - Wang, Hong

AU - Jiang, Zhong Ping

PY - 2024

Y1 - 2024

N2 - This article proposes a deep learning (DL)-based control algorithm - DL velocity-based model predictive control (VMPC) - for reducing traffic congestion with slowly time-varying traffic signal controls. This control algorithm consists of system identification using DL and traffic signal control using VMPC. For the training process of DL, we established a modeling error entropy loss as the criteria inspired by the theory of stochastic distribution control (SDC) originated by the fourth author. Simulation results show that the proposed algorithm can reduce traffic congestion with a slowly varying traffic signal control input. Results of an ablation study demonstrate that this algorithm compares favorably to other model-based controllers in terms of prediction error, signal varying speed, and control effectiveness.

AB - This article proposes a deep learning (DL)-based control algorithm - DL velocity-based model predictive control (VMPC) - for reducing traffic congestion with slowly time-varying traffic signal controls. This control algorithm consists of system identification using DL and traffic signal control using VMPC. For the training process of DL, we established a modeling error entropy loss as the criteria inspired by the theory of stochastic distribution control (SDC) originated by the fourth author. Simulation results show that the proposed algorithm can reduce traffic congestion with a slowly varying traffic signal control input. Results of an ablation study demonstrate that this algorithm compares favorably to other model-based controllers in terms of prediction error, signal varying speed, and control effectiveness.

KW - Deep learning (DL)

KW - gain-scheduling

KW - model predictive control (MPC)

KW - traffic signal control

KW - velocity-based linearization (VL)

UR - http://www.scopus.com/inward/record.url?scp=85153488721&partnerID=8YFLogxK

U2 - 10.1109/TNNLS.2023.3264709

DO - 10.1109/TNNLS.2023.3264709

M3 - Article

C2 - 37053060

AN - SCOPUS:85153488721

SN - 2162-237X

VL - 35

SP - 12760

EP - 12771

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

IS - 9

ER -

Reducing Urban Traffic Congestion Using Deep Learning and Model Predictive Control

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this