PressLight

Wei, Hua; Chen, Chacha; Zheng, Guopei; Wu, Kan; Gayah, Vikash V.; Xu, Kai; Li, Zhenhui

doi:10.1145/3292500.3330949

Cited by 175 publications

(43 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We have applied five methods, including Fixed-time method [20], Q-learning (Center) [46], Q-learning (Edge) [46], Nash Q-learning [47], and our MAAC method. They were all trained in 1000 episodes in CityFLow [44] based on edge computing architecture as we designed. As shown in Figure 12, we can see that the the algorithms converged at around 600 episode point.…”

Section: Resultsmentioning

confidence: 99%

“…We apply an open source simulator for traffic environment: CityFlow [44] as our experiment environment. We assumed that there are six traffic lights (intersection nodes or edge computing nodes) in one section of a city (as shown in Figure 11).…”

Section: Simulation Environment and Settingsmentioning

confidence: 99%

“…We assumed that there are six traffic lights (intersection nodes or edge computing nodes) in one section of a city (as shown in Figure 11). Our dynamic control of the traffic lights was using the CityFlow [44] Python interface at runtime. Here are the settings of our experiments (as shown in Table 2).…”

Section: Simulation Environment and Settingsmentioning

confidence: 99%

See 2 more Smart Citations

An Edge Based Multi-Agent Auto Communication Method for Traffic Light Control

Yong

et al. 2020

Sensors

View full text Add to dashboard Cite

With smart city infrastructures growing, the Internet of Things (IoT) has been widely used in the intelligent transportation systems (ITS). The traditional adaptive traffic signal control method based on reinforcement learning (RL) has expanded from one intersection to multiple intersections. In this paper, we propose a multi-agent auto communication (MAAC) algorithm, which is an innovative adaptive global traffic light control method based on multi-agent reinforcement learning (MARL) and an auto communication protocol in edge computing architecture. The MAAC algorithm combines multi-agent auto communication protocol with MARL, allowing an agent to communicate the learned strategies with others for achieving global optimization in traffic signal control. In addition, we present a practicable edge computing architecture for industrial deployment on IoT, considering the limitations of the capabilities of network transmission bandwidth. We demonstrate that our algorithm outperforms other methods over 17% in experiments in a real traffic simulation environment.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Simulation Environment and Settingsmentioning

confidence: 99%

See 1 more Smart Citation

An Edge Based Multi-Agent Auto Communication Method for Traffic Light Control

Yong

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…The RL-based traffic signal control methods can be divided into three categories depending on its control areas: single intersection traffic signal control, arterial traffic signal control, and network traffic signal control. The RL-based methods were reported to have better performance than fixed-time and actuated methods regardless of their control areas [19][20][21][22]. However, these studies consider only unimodal traffic in reward functions, limiting their capabilities to model heterogeneous interests, and complex interactions of different traffic modes at intersections.…”

Section: Related Workmentioning

confidence: 99%

Human‐centric multimodal deep (HMD) traffic signal control

Wang

Dong

et al. 2022

IET Intelligent Trans Sys

View full text Add to dashboard Cite

Reinforcement learning (RL)‐based models have been widely studied for traffic signal control with objectives, such as minimizing vehicle delay and queue length, maximizing vehicle throughput, and improving road safety, through tailored reward designs. Despite the advancements in RL‐based signal control models for car traffic, limited research focused on multimodal traffic (e.g. car, bus, pedestrian). The simplified environment of unimodal traffic restrains these models from applications in real‐world cases. In this paper, the authors propose an RL‐based human‐centric multimodal deep (HMD) traffic signal control method to coordinate multimodal traffic at an intersection, with the objective of minimizing the waiting time per capita for multiple traffic modes by taking consideration of the mode capacity to ensure social equity. The method is validated in the simulation of urban mobility (SUMO) simulation environment using real traffic data. The results show the superior performance of HMD over vehicle‐centric (VC) RL‐based methods and traditional signal control schemes. HMD reduces the waiting time per capita by up to 19.2% and 6.3% compared with the fixed timing and VC RL‐based methods. In addition, the experiment results show that the HMD method assigns a higher priority to public transport over low‐occupancy travel modes in passing through intersections compared with unimodal traffic signal control strategies.

show abstract

“…However, different local optima in the policy space can correspond to strategies that differ in nature, which makes the above consensus problematic in RL tasks where the environment is unstable. For example, in adaptive traffic signal control (ATSC) [81,89,90] (conceptual diagram and more examples are included in Figure 1), if two traffic flows are desired to reach the target points from the departure points quickly, multiple control strategies with similar average commuting times may exist due to the combinatorial nature of traffic lights. The performance of a single policy obtained by reward maximization is bound to be affected if the subsequent traffic volumes on other sections of the road network associated with the traveled section of that traffic change.…”

Section: Introductionmentioning

confidence: 99%

Diverse Policy Optimization for Structured Action Space

Li¹,

Wang²,

Yang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Enhancing the diversity of policies is beneficial for robustness, exploration, and transfer in reinforcement learning (RL). In this paper, we aim to seek diverse policies in an under-explored setting, namely RL tasks with structured action spaces with the two properties of composability and local dependencies. The complex action structure, non-uniform reward landscape, and subtle hyperparameter tuning due to the properties of structured actions prevent existing approaches from scaling well. We propose a simple and effective RL method, Diverse Policy Optimization (DPO), to model the policies in structured action space as the energy-based models (EBM) by following the probabilistic RL framework. A recently proposed novel and powerful generative model, GFlowNet, is introduced as the efficient, diverse EBM-based policy sampler. DPO follows a joint optimization framework: the outer layer uses the diverse policies sampled by the GFlowNet to update the EBM-based policies, which supports the GFlowNet training in the inner layer. Experiments on ATSC and Battle benchmarks demonstrate that DPO can efficiently discover surprisingly diverse policies in challenging scenarios and substantially outperform existing state-of-the-art methods.

show abstract

PressLight

Cited by 175 publications

References 15 publications

An Edge Based Multi-Agent Auto Communication Method for Traffic Light Control

An Edge Based Multi-Agent Auto Communication Method for Traffic Light Control

Human‐centric multimodal deep (HMD) traffic signal control

Diverse Policy Optimization for Structured Action Space

Contact Info

Product

Resources

About