Optimizing Large-Scale Fleet Management on a Road Network using Multi-Agent Deep Reinforcement Learning with Graph Neural Network

被引:10
作者
Kim, Juhyeon [1 ]
Kim, Kihyun [1 ]
机构
[1] Seoul Natl Univ, Dept Elect & Comp Engn, Seoul 08826, South Korea
来源
2021 IEEE INTELLIGENT TRANSPORTATION SYSTEMS CONFERENCE (ITSC) | 2021年
关键词
D O I
10.1109/ITSC48978.2021.9565029
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We propose a novel approach to optimize fleet management by combining multi-agent reinforcement learning with graph neural network. To provide ride-hailing service, one needs to optimize dynamic resources and demands over spatial domain. While the spatial structure was previously approximated with a regular grid, our approach represents the road network with a graph, which better reflects the underlying geometric structure. Dynamic resource allocation is formulated as multi-agent reinforcement learning, whose action-value function (Q function) is approximated with graph neural networks. We use stochastic policy update rule over the graph with deep Q-networks (DQN), and achieve superior results over the greedy policy update. We design a realistic simulator that emulates the empirical taxi call data, and confirm the effectiveness of the proposed model under various conditions.
引用
收藏
页码:990 / 995
页数:6
相关论文
共 25 条
[1]  
Alshamsi A., 2009, P 8 INT C AUT AG MUL, P21
[2]  
Bello I, 2017, arXiv, DOI [DOI 10.48550/ARXIV.1611.09940, 10.48550/arxiv.1611.09940]
[3]  
Dai HJ, 2017, ADV NEUR IN, V30
[4]  
Foerster JN, 2018, AAAI CONF ARTIF INTE, P2974
[5]   An adaptive dynamic programming algorithm for dynamic fleet management, I: Single period travel times [J].
Godfrey, GA ;
Powell, WB .
TRANSPORTATION SCIENCE, 2002, 36 (01) :21-39
[6]  
Haarnoja T, 2017, PR MACH LEARN RES, V70
[7]  
Ke J., 2019, arXiv
[8]   Taxi dispatch system based on current demands and real-time traffic conditions [J].
Lee, DH ;
Wang, H ;
Cheu, RL ;
Teo, SH .
TRANSPORTATION NETWORK MODELING 2004, 2004, (1882) :193-200
[9]   Efficient Ridesharing Order Dispatching with Mean Field Multi-Agent Reinforcement Learning [J].
Li, Minne ;
Qin, Zhiwei ;
Jiao, Yan ;
Yang, Yaodong ;
Gong, Zhichen ;
Wang, Jun ;
Wang, Chenxi ;
Wu, Guobin ;
Ye, Jieping .
WEB CONFERENCE 2019: PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE (WWW 2019), 2019, :983-994
[10]  
Li XH, 2019, AAMAS '19: PROCEEDINGS OF THE 18TH INTERNATIONAL CONFERENCE ON AUTONOMOUS AGENTS AND MULTIAGENT SYSTEMS, P980