Yulin Tian, Shuqing Liu, Lu Wei, Zhen Li, Shaohu Tang, Yuchen Zhang, Tao Zhu
{"title":"Traffic signal active control method for short-distance intersections.","authors":"Yulin Tian, Shuqing Liu, Lu Wei, Zhen Li, Shaohu Tang, Yuchen Zhang, Tao Zhu","doi":"10.1371/journal.pone.0319804","DOIUrl":null,"url":null,"abstract":"<p><p>Aiming at the existing problems about the overflow prevention goal and the overall traffic efficiency guarantee being difficult to optimize at the same time in the signal control process of short-distance intersections scenario, this paper proposes a traffic signal active control method based on key state prediction. In order to construct the key state evolution trend of short-distance intersection scenarios, this paper proposes the concept of overflow index for short-distance road sections and designs the prediction method of overflow index. In order to perform fast computation and solution for the active control scheme, this paper builds a solution algorithm based on deep reinforcement learning and optimizes the problem of reward sparsity in the algorithm, which improves the ability of active control in terms of state space and reward function. The experimental results show that this method can not only ensure the overall traffic efficiency of short-distance intersections and reduce the travel delay but also can actively sense the change of overflow state, improve the overflow prevention and control ability of the target scenario, and reduce the overflow risk.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 3","pages":"e0319804"},"PeriodicalIF":2.9000,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"PLoS ONE","FirstCategoryId":"103","ListUrlMain":"https://doi.org/10.1371/journal.pone.0319804","RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/1/1 0:00:00","PubModel":"eCollection","JCR":"Q1","JCRName":"MULTIDISCIPLINARY SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
Aiming at the existing problems about the overflow prevention goal and the overall traffic efficiency guarantee being difficult to optimize at the same time in the signal control process of short-distance intersections scenario, this paper proposes a traffic signal active control method based on key state prediction. In order to construct the key state evolution trend of short-distance intersection scenarios, this paper proposes the concept of overflow index for short-distance road sections and designs the prediction method of overflow index. In order to perform fast computation and solution for the active control scheme, this paper builds a solution algorithm based on deep reinforcement learning and optimizes the problem of reward sparsity in the algorithm, which improves the ability of active control in terms of state space and reward function. The experimental results show that this method can not only ensure the overall traffic efficiency of short-distance intersections and reduce the travel delay but also can actively sense the change of overflow state, improve the overflow prevention and control ability of the target scenario, and reduce the overflow risk.
期刊介绍:
PLOS ONE is an international, peer-reviewed, open-access, online publication. PLOS ONE welcomes reports on primary research from any scientific discipline. It provides:
* Open-access—freely accessible online, authors retain copyright
* Fast publication times
* Peer review by expert, practicing researchers
* Post-publication tools to indicate quality and impact
* Community-based dialogue on articles
* Worldwide media coverage