AI 中文总结
针对网联车辆路径规划的双目标挑战，提出AGDRP方法，采用DRL平衡行驶时间与AoI，仿真显示其性能优于仅优化行驶时间的基准方案。
AI 中文摘要
智能交通系统（ITS）的发展很大程度上得益于无线通信技术的进步。动态路径规划作为ITS的关键组成部分，传统上聚焦于路径容量、行驶时间等指标。本文提出一种新颖的双因素方法，将行驶时间估计与无线资源可用性相结合，形成面向网联车辆（CVs）的创新路径规划方案。为解决这一双目标路径规划挑战，采用深度强化学习（DRL），所提出的方法命名为保障信息年龄的动态路径规划（AGDRP），可有效平衡行驶时间与信息年龄（AoI），通过随时间的自适应学习提升路径规划性能。仿真结果表明，AGDRP优于仅聚焦行驶时间优化的基准方案，证实纳入AoI最小化可显著提升路径规划性能，超越传统基于行驶时间的方法。
英文摘要
The advancement of Intelligent Transportation Sys- tems (ITS) has been significantly driven by progress in radio communication technology. Dynamic route planning, a key com- ponent of ITS, traditionally focuses on metrics such as route capacity and travel time. This paper presents a novel dual- factor approach that integrates travel time estimation and radio resource availability into an innovative route-planning scheme for connected vehicles (CVs). To address this dual-objective route planning challenge, we employ Deep Reinforcement Learning (DRL). Our approach, called AoI-Guaranteed Dynamic Route Planning (AGDRP), effectively balances travel time and Age of Information (AoI), enhancing route planning performance through adaptive learning over time. Simulation results demon- strate that AGDRP outperforms the baseline scheme, which solely focuses on travel time optimization. In fact, we show that incor- porating AoI minimization significantly enhances route planning performance beyond conventional travel-time-based approaches.
