Scholarlink

<p>To fully utilize the computational resources in the cooperative control environment and achieve global optimization for connected and automated vehicles, a parallel distributed computing framework is presented by a decomposition strategy. This strategy converts the original centralized optimization problem into a separable form by introducing a set of auxiliary variables and consensus equality constraints to address the coupling components necessary for collision avoidance. Based on the numerical analysis of communication resource consumption in this distributed framework, an information filtering strategy is further designed to enhance communication efficiency by limiting the transmission of consensus variables. Consequently, the global convergence of this parallel algorithm with filtered information is theoretically analyzed under the assumption that the feasible domain is convex. The communication traffic, cooperative control, numerical optimization performance of the proposed framework and parallel algorithm are validated and assessed through several simulation and experimental tests with the intersection scenario.</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 Parallel Distributed Framework for Cooperative Control of CAVs with Less Communication Load

  • 期刊论文

  • 发表日期：

    2025年6月9日

查看全文
  • Jie Ma,
  • Qiuxia Hu,
  • Feng Gao,
  • Guanglun Zhan

摘要

To fully utilize the computational resources in the cooperative control environment and achieve global optimization for connected and automated vehicles, a parallel distributed computing framework is presented by a decomposition strategy. This strategy converts the original centralized optimization problem into a separable form by introducing a set of auxiliary variables and consensus equality constraints to address the coupling components necessary for collision avoidance. Based on the numerical analysis of communication resource consumption in this distributed framework, an information filtering strategy is further designed to enhance communication efficiency by limiting the transmission of consensus variables. Consequently, the global convergence of this parallel algorithm with filtered information is theoretically analyzed under the assumption that the feasible domain is convex. The communication traffic, cooperative control, numerical optimization performance of the proposed framework and parallel algorithm are validated and assessed through several simulation and experimental tests with the intersection scenario.