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:1109.5319

再生可能資源の最適採餌戦略

Optimal Foraging of Renewable Resources

シェア:XThreadsFacebookLINEはてブBluesky

確率的に出現する目標点を限られたセンサ範囲で探索・訪問するエージェント群の待ち時間を最小化する最適方策を導出し、探索支配時は出現率の平方根、処理支配時は立方根に比例して領域を巡回すべきことを示した。

著者: John J. Enright, Emilio Frazzoli

分類: cs.RO

原文アブストラクト

Consider a team of agents in the plane searching for and visiting target points that appear in a bounded environment according to a stochastic renewal process with a known absolutely continuous spatial distribution. Agents must detect targets with limited-range onboard sensors. It is desired to minimize the expected waiting time between the appearance of a target point, and the instant it is visited. When the sensing radius is small, the system time is dominated by time spent searching, and it is shown that the optimal policy requires the agents to search a region at a relative frequency proportional to the square root of its renewal rate. On the other hand, when targets appear frequently, the system time is dominated by time spent servicing known targets, and it is shown that the optimal policy requires the agents to service a region at a relative frequency proportional to the cube root of its renewal rate. Furthermore, the presented algorithms in this case recover the optimal performance achieved by agents with full information of the environment. Simulation results verify the theoretical performance of the algorithms.

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

FRONTIER(GPU搭載BTOパソコン)

FRONTIER(GPU搭載BTOパソコン)の公式広告バナーFRONTIERでGPU搭載PCを見る ↗
PR本紙発行元 EmplifAI