日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

群制御arXiv:2106.11763

車線選好を考慮した自動運転車の隊列制御

Formation Control with Lane Preference for Connected and Automated Vehicles in Multi-lane Scenarios

シェア:XThreadsFacebookLINEはてブBluesky

多車線道路で各車の希望車線を考慮し、二層型隊列制御とCBSで衝突回避経路を計画する手法を提案し、高交通量時に渋滞を減らし効率を改善することを示した。

著者: Mengchi Cai, Chaoyi Chen, Jiawei Wang, Qing Xu, Keqiang Li, Jianqiang Wang, Xiangbin Wu

分類: cs.RO, cs.SY, eess.SY

原文アブストラクト

Multi-lane roads are typical scenarios in the real-world traffic system. Vehicles usually have preference on lanes according to their routes and destinations. Few of the existing studies looks into the problem of controlling vehicles to drive on their desired lanes. This paper proposes a formation control method that considers vehicles' preference on different lanes. The bi-level formation control framework is utilized to plan collision-free motion for vehicles, where relative target assignment and path planning are performed in the upper level, and trajectory planning and tracking are performed in the lower level. The collision-free multi-vehicle path planning problem considering lane preference is decoupled into two sub problems: calculating assignment list with non-decreasing cost and planning collision-free paths according to given assignment result. The Conflict-based Searching (CBS) method is utilized to plan collision-free paths for vehicles based on given assignment results. Case study is conducted and simulations are carried out in a three-lane road scenario. The results indicate that the proposed formation control method significantly reduces congestion and improves traffic efficiency at high traffic volumes, compared to the rule-based method.

関連論文

  1. 1Boston Dynamics、車輪付きAtlasの研究試作機を紹介研究・技術
  2. 2newmoとKDDIが資本業務提携 2028年に堺市で自動運転タクシー商用化目指す提携・買収
  3. 3GXOとApptronik、Apolloの物流現場試験を拡大 最大1,000台の展開を構想量産・展開
  4. 4NVIDIA、CoRLでGR00T 2とIsaac Lab 3.0を発表新製品・発表
  5. 5UC BerkeleyとNVIDIAの研究者、AIエージェントでロボット制御コードを仮想環境で改善研究・技術

PR・アフィリエイト広告

MINISFORUM(ミニPC)

MINISFORUM(ミニPC)の公式広告バナーMINISFORUM公式ストアを見る ↗
PR本紙発行元 EmplifAI