Efficient Large-Scale Multi-Drone Delivery Using Transit Networks

被引:20
|
作者
Choudhury, Shushman [1 ]
Solovey, Kiril [1 ]
Kochenderfer, Mykel J. [1 ]
Pavone, Marco [1 ]
机构
[1] Stanford Univ, Stanford, CA 94305 USA
基金
美国国家科学基金会;
关键词
TRAVELING SALESMAN PROBLEM; OPTIMIZATION; ALGORITHMS;
D O I
10.1109/icra40945.2020.9197313
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We consider the problem of controlling a large fleet of drones to deliver packages simultaneously across broad urban areas. To conserve energy, drones hop between public transit vehicles (e.g., buses and trams). We design a comprehensive algorithmic framework that strives to minimize the maximum time to complete any delivery. We address the multifaceted complexity of the problem through a two-layer approach. First, the upper layer assigns drones to package delivery sequences with a near-optimal polynomial-time task allocation algorithm. Then, the lower layer executes the allocation by periodically routing the fleet over the transit network while employing efficient bounded-suboptimal multi-agent pathfinding techniques tailored to our setting. Experiments demonstrate the efficiency of our approach on settings with up to 200 drones, 5000 packages, and transit networks with up to 8000 stops in San Francisco and Washington DC. Our results show that the framework computes solutions within a few seconds (up to 2 minutes at most) on commodity hardware, and that drones travel up to 450% of their flight range with public transit.
引用
收藏
页码:4543 / 4550
页数:8
相关论文
共 50 条
  • [21] Opportunistic Multi-Drone Networks: Filling the Spatiotemporal Holes of Collaborative and Distributed Applications
    Flores H.
    IEEE Inter. Things Magazine, 2024, 2 (94-100): : 94 - 100
  • [22] Automatic multi-drone path planning using network flow algorithms
    Seo Y.
    Lee H.
    Na H.-S.
    Journal of Institute of Control, Robotics and Systems, 2019, 25 (04) : 303 - 311
  • [23] Dynamic Replanning of Multi-drone Missions using Dynamic Forward Slicing
    Campusano, Miguel
    Lundquist, Ulrik Pagh Schultz
    PROCEEDINGS OF THE 21ST ACM SIGPLAN INTERNATIONAL CONFERENCE ON GENERATIVE PROGRAMMING: CONCEPTS AND EXPERIENCES, GPCE 2022, 2022, : 72 - 85
  • [24] Efficient Training of Large-Scale Neural Networks Using Linear Pipeline Broadcast
    Yu, Chanhee
    Park, Kyongseok
    IEEE ACCESS, 2024, 12 : 165653 - 165662
  • [25] Efficient synthesis of large-scale heat exchanger networks using monogenetic algorithm
    Ernst, Philipp
    Fieg, Georg
    Luo, Xing
    HEAT AND MASS TRANSFER, 2010, 46 (10) : 1087 - 1096
  • [26] Efficient Sampling Strategies for Large-scale Complex Networks
    Yang Bo
    Gao Hai-xia
    Chen Zhong
    2008 INTERNATIONAL CONFERENCE ON MANAGEMENT SCIENCE & ENGINEERING (15TH), VOLS I AND II, CONFERENCE PROCEEDINGS, 2008, : 334 - +
  • [27] Efficient parallel simulation of large-scale PCS networks
    Boukerche, A
    Das, SK
    Fabbri, A
    Yildiz, O
    TRANSACTIONS OF THE SOCIETY FOR COMPUTER SIMULATION INTERNATIONAL, 1999, 16 (03): : 113 - 125
  • [28] Efficient synthesis of large-scale heat exchanger networks using monogenetic algorithm
    Philipp Ernst
    Georg Fieg
    Xing Luo
    Heat and Mass Transfer, 2010, 46 : 1087 - 1096
  • [29] Efficient Targeting of Sensor Networks for Large-Scale Systems
    Choi, Han-Lim
    How, Jonathan P.
    IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, 2011, 19 (06) : 1569 - 1577
  • [30] Efficient localization for large-scale underwater sensor networks
    Zhou, Zhong
    Cui, Jun-Hong
    Zhou, Shengli
    AD HOC NETWORKS, 2010, 8 (03) : 267 - 279