Deep Reinforcement Learning based Traffic Signal optimization for Multiple Intersections in ITS

Paul, Ananya; Mitra, Sulata

doi:10.1109/ants50601.2020.9342819

Cited by 12 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The current policy at tth time can be considered as long as this policy is not too dissimilar from the previous policy of ðt À 1Þth time. The policy ratio rt t is defined as (3) to measure the difference between the current and previous policies.…”

Section: Ppomentioning

confidence: 99%

See 1 more Smart Citation

Exploring reward efficacy in traffic management using deep reinforcement learning in intelligent transportation system

Paul¹,

Mitra²

2022

ETRI Journal

Self Cite

View full text Add to dashboard Cite

In the last decade, substantial progress has been achieved in intelligent traffic control technologies to overcome consistent difficulties of traffic congestion and its adverse effect on smart cities. Edge computing is one such advanced progress facilitating real-time data transmission among vehicles and roadside units to mitigate congestion. An edge computing-based deep reinforcement learning system is demonstrated in this study that appropriately designs a multiobjective reward function for optimizing different objectives. The system seeks to overcome the challenge of evaluating actions with a simple numerical reward. The selection of reward functions has a significant impact on agents' ability to acquire the ideal behavior for managing multiple traffic signals in a large-scale road network. To ascertain effective reward functions, the agent is trained withusing the proximal policy optimization method in several deep neural network models, including the state-of-the-art transformer network. The system is verified using both hypothetical scenarios and real-world traffic maps. The comprehensive simulation outcomes demonstrate the potency of the suggested reward functions.

show abstract

Section: Ppomentioning

confidence: 99%

“…Conventional RL algorithms are generally ineffective in large‐scale complex environments due to combinatorial explosion [3]. To address this issue, a deep neural network (DNN), a function approximator, is integrated with RL to create deep RL (DRL).…”

Section: Introductionmentioning

confidence: 99%

Exploring reward efficacy in traffic management using deep reinforcement learning in intelligent transportation system

Paul¹,

Mitra²

2022

ETRI Journal

Self Cite

View full text Add to dashboard Cite

show abstract

“…State: State representation should contain ample amounts of traffic information to reflect the dynamic nature of the traffic. 25 This information can be gathered by diverse methods, such as sensor data retrieved from vehicles and traffic signals.…”

Section: Mdp Settingmentioning

confidence: 99%

Deep reinforcement learning based cooperative control of traffic signal for multi‐intersection network in intelligent transportation system using edge computing

Paul

2022

Trans Emerging Tel Tech

Self Cite

View full text Add to dashboard Cite

In the current era, the coordination of traffic flow is hindered by the discrepancy between road infrastructure and the number of vehicles which leads to traffic congestion. One of the widely used strategies to mitigate traffic congestion is to control traffic signals with the help of deep reinforcement learning (DRL) in edge computing based intelligent transportation system. This article provides a comprehensive analysis of the most recent DRL algorithms, advantage actor‐critic and proximal policy optimization in multiple deep neural networks (DNNs), including a state‐of‐the‐art transformer model for effective traffic signal management. Here, a single DRL agent is used, which obtains the spatio‐temporal information of the traffic to identify traffic patterns from complex intersection environments. The agent uses this information as the input to the DNNs and then applies the algorithms to retrieve the essential parameters of DNN to seek an optimal action selection policy to mitigate congestion. Different real‐time maps and small city networks are explored here to determine which DNN is best suited for traffic congestion management. The simulation study reveals that both the algorithms significantly outperform the baseline. The transformer model gives the best result when compared to other DNNs. The transformer model decreases average waiting time by 96.16%, implying that it has a higher capability of dealing with congested environments.

show abstract

“…They mainly rely on optimisation problems for the traffic light phase. Many recent solutions are based on deep reinforcement learning techniques [15][16][17][18][19][20][21] or again ARIMA processes [22].…”

Section: Introductionmentioning

confidence: 99%

Signalized and Unsignalized Road Traffic Intersection Models: A Comprehensive Benchmark Analysis

Ba,

Tordeux

2023

Coll Dyn

View full text Add to dashboard Cite

Road traffic flow models allow the development and testing of intelligent transportation solutions. Macroscopic intersection models are especially relevant for the simulation of large traffic networks. In this article, we study four first-order signalized and unsignalized intersection models. The two unsignalized approaches are the first-in-first-out (FIFO) model (roundabout-type intersection) and an optimal non-FIFO model (highway-type intersection). The optimal control operates upstream for the first signalized intersection model. It occurs downstream for the second signalized model. All four models satisfy the expected physical constraints of vehicle conservation, traffic demand, and assignment. The models are minimal and allow a comprehensible analysis of the results. We determine mathematical relationships between the intersection models and empirically analyze the performances using Monte Carlo simulations. The numerical simulations assume random demand, supply, and assignment. Besides average performances, the approach accounts for the flow ranges of variation. A benchmark analysis compares the intersection models. We observe that the optimal signalized intersection models overcome the performances of the FIFO model in congested states. They may even reach the performances of the idealistic non-FIFO model. Further applications for the four intersection models are discussed.

show abstract

Deep Reinforcement Learning based Traffic Signal optimization for Multiple Intersections in ITS

Cited by 12 publications

References 9 publications

Exploring reward efficacy in traffic management using deep reinforcement learning in intelligent transportation system

Exploring reward efficacy in traffic management using deep reinforcement learning in intelligent transportation system

Deep reinforcement learning based cooperative control of traffic signal for multi‐intersection network in intelligent transportation system using edge computing

Signalized and Unsignalized Road Traffic Intersection Models: A Comprehensive Benchmark Analysis

Contact Info

Product

Resources

About