Scholarlink

<p>Vehicle-road cooperative perception plays an important role in intelligent transportation systems. By receiving cooperative perception messages (CPM) shared from the roadside infrastructure, the vehicle can acquire richer real-time information about road conditions and traffic flow. However, existing CPM sharing strategies face challenges when dealing with complex traffic scenarios, including data exchange frequency leading to channel congestion, and indiscriminate data sharing resulting in redundant or incomplete information. In this regard, this paper proposes a cybertwin-based cooperative perception sharing scheme (CPCS). Leveraging the perceptual information collected by cybertwin, we design a road individual intention relevance criterion (RIRC) to quantify the traffic safety relationship among road individuals and propose a high-value content selection learning algorithm based on deep deterministic policy gradient and velocity entropy. By calculating the relevance between individual road users, the infrastructure selectively shares road information with vehicles, thus improving the timeliness and effectiveness of CPM. Furthermore, considering the diversity of vehicle perception demands and the fluctuation of real-time channel busy ratio, we design a CPM adaptive transmission frequency control and bandwidth allocation algorithm under multi-vehicle requests to avoid channel congestion and resource wastage. Compared to traditional sharing schemes, the CPM quality of service is improved by 2.45 times, and the object redundancy is reduced by 75.4%.</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

CPCS: a perception sharing scheme of vehicle-road cooperation based on cybertwin

  • 期刊论文

  • 发表日期：

    2025年1月24日

查看全文
  • Jianhang Liu,
  • Chunxing Xia,
  • Xuerong Cui,
  • Haibo Wu

摘要

Vehicle-road cooperative perception plays an important role in intelligent transportation systems. By receiving cooperative perception messages (CPM) shared from the roadside infrastructure, the vehicle can acquire richer real-time information about road conditions and traffic flow. However, existing CPM sharing strategies face challenges when dealing with complex traffic scenarios, including data exchange frequency leading to channel congestion, and indiscriminate data sharing resulting in redundant or incomplete information. In this regard, this paper proposes a cybertwin-based cooperative perception sharing scheme (CPCS). Leveraging the perceptual information collected by cybertwin, we design a road individual intention relevance criterion (RIRC) to quantify the traffic safety relationship among road individuals and propose a high-value content selection learning algorithm based on deep deterministic policy gradient and velocity entropy. By calculating the relevance between individual road users, the infrastructure selectively shares road information with vehicles, thus improving the timeliness and effectiveness of CPM. Furthermore, considering the diversity of vehicle perception demands and the fluctuation of real-time channel busy ratio, we design a CPM adaptive transmission frequency control and bandwidth allocation algorithm under multi-vehicle requests to avoid channel congestion and resource wastage. Compared to traditional sharing schemes, the CPM quality of service is improved by 2.45 times, and the object redundancy is reduced by 75.4%.