Scholarlink

<p>Adaptive Traffic Signal Control (ATSC) has emerged as a core component of intelligent urban mobility. Yet its transition from simulation to reliable real-world operation remains persistently constrained. This survey synthesises 211 peer-reviewed publications spanning January 2017 to early March 2026 around five specific contributions. First, a unified taxonomy classifies ATSC systems across control architectures, methodological paradigms, and application contexts, enabling systematic cross-paradigm comparison absent from prior surveys. Second, a comparative analysis across eight operational dimensions establishes that the inverse relationship between algorithmic sophistication and deployment readiness is a structural feature of the field. Third, a substantive review of real-world deployed systems reveals that operational success reflects graceful degradation and conservative adaptation rather than algorithmic sophistication. Fourth, a dedicated review of large language model applications in ATSC covers zero-shot control, hybrid reinforcement learning architectures, cooperative multi-agent frameworks, and live field deployments. Six structural limitations defining the agenda for operational deployment are identified. Fifth, the evaluation framework is extended to incorporate deployment readiness, interpretability, multimodal support, and critical gap assessments across all reviewed dimensions. The survey identifies vehicle-centrism as a governance failure embedded in the research agenda. Simulation dominance is established as universal across all reviewed technology categories. A dependency-structured research agenda is proposed across seven directions: robust generalizable learning, federated and edge AI, multimodal sensor fusion, explainability and human-in-the-loop control, equity-oriented optimization, socio-technical integration, and LLM-augmented context-aware control. This survey provides a deployment-oriented reference for researchers, practitioners, and policymakers advancing equitable and resilient urban traffic signal control.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Comprehensive Survey of Adaptive Traffic Signal Control: Methods, Applications, Challenges, and Future Research

  • 期刊论文

  • 发表日期：

    2026年4月14日

查看全文
  • Mahbub Hassan,
  • Turjoy Das Turjo,
  • Ali Hanafiah Rambe,
  • Muzammil Jusoh,
  • Md Kamrul Islam,
  • M. M. Hafizur Rahman,
  • Md Shafiul Alam,
  • Muhammad Muhitur Rahman

摘要

Adaptive Traffic Signal Control (ATSC) has emerged as a core component of intelligent urban mobility. Yet its transition from simulation to reliable real-world operation remains persistently constrained. This survey synthesises 211 peer-reviewed publications spanning January 2017 to early March 2026 around five specific contributions. First, a unified taxonomy classifies ATSC systems across control architectures, methodological paradigms, and application contexts, enabling systematic cross-paradigm comparison absent from prior surveys. Second, a comparative analysis across eight operational dimensions establishes that the inverse relationship between algorithmic sophistication and deployment readiness is a structural feature of the field. Third, a substantive review of real-world deployed systems reveals that operational success reflects graceful degradation and conservative adaptation rather than algorithmic sophistication. Fourth, a dedicated review of large language model applications in ATSC covers zero-shot control, hybrid reinforcement learning architectures, cooperative multi-agent frameworks, and live field deployments. Six structural limitations defining the agenda for operational deployment are identified. Fifth, the evaluation framework is extended to incorporate deployment readiness, interpretability, multimodal support, and critical gap assessments across all reviewed dimensions. The survey identifies vehicle-centrism as a governance failure embedded in the research agenda. Simulation dominance is established as universal across all reviewed technology categories. A dependency-structured research agenda is proposed across seven directions: robust generalizable learning, federated and edge AI, multimodal sensor fusion, explainability and human-in-the-loop control, equity-oriented optimization, socio-technical integration, and LLM-augmented context-aware control. This survey provides a deployment-oriented reference for researchers, practitioners, and policymakers advancing equitable and resilient urban traffic signal control.