Efficient Ridesharing Order Dispatching with Mean Field Multi-Agent Reinforcement Learning
Open accessThe DOI matches its Crossref record; 39 of the 39 references it cites were found in a registry or an index.
Abstract
A fundamental question in any peer-to-peer ridesharing system is how to, both effectively and efficiently, dispatch user's ride requests to the right driver in real time. Traditional rule-based solutions usually work on a simplified problem setting, which requires a sophisticated hand-crafted weight design for either centralized authority control or decentralized multi-agent scheduling systems. Although recent approaches have used reinforcement learning to provide centralized combinatorial optimization algorithms with informative weight values, their single-agent setting can hardly model the complex interactions between drivers and orders. In this paper, we address the order dispatching problem using multi-agent reinforcement learning (MARL), which follows the distributed nature of the peer-to-peer ridesharing problem and possesses the ability to capture the stochastic demand-supply dynamics in large-scale ridesharing scenarios. Being more reliable than centralized approaches, our proposed MARL solutions could also support fully distributed execution through recent advances in the Internet of Vehicles (IoV) and the Vehicle-to-Network (V2N). Furthermore, we adopt the mean field approximation to simplify the local interactions by taking an average action among neighborhoods. The mean field approximation is capable of globally capturing dynamic demand-supply variations by propagating many local interactions between agents and the environment. Our extensive experiments have shown the significant improvements of MARL order dispatching algorithms over several strong baselines on the accumulated driver income (ADI), and order response rate measures. Besides, the simulated experiments with real data have also justified that our solution can alleviate the supply-demand gap during the rush hours, thus possessing the capability of reducing traffic congestion.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357 · cited 2×
“ther than a given agent i with the subscript −i, e.g., a ≡(at −i ,a t i ). To stabilize the training process, we maintain an experience replay bufferDcontaining tuples (ot ,at rt,ot+1)as described in [26]. 2.1.2 Dynamic of the Action Set Elements. In the order dispatching problem, an action is defined as an active ord…”
MethodInfluentialThe Internet of Things: A survey
Atzori, L., Iera, A., & Morabito, G. (2010) · Computer Networks · doi:10.1016/j.comnet.2010.05.010 · Cited by 15,613 · cited 1×
“With the recent development of the Internet of Things (IoT) [5, 11, 43] and the Internet of Vehicles (IoV) [22, 41], the fully distributed execution could be practically deployed by distributing the centralized trained policy to each vehicle through the Vehicle-to-Network (V2N) [1].”
(2010) · International Conference on Machine Learning · Cited by 13,995
Internet of Things (IoT): A vision, architectural elements, and future directions
Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M. (2013) · Future Generation Computer Systems · doi:10.1016/j.future.2013.01.010 · Cited by 12,293 · cited 1×
“With the recent development of the Internet of Things (IoT) [5, 11, 43] and the Internet of Vehicles (IoV) [22, 41], the fully distributed execution could be practically deployed by distributing the centralized trained policy to each vehicle through the Vehicle-to-Network (V2N) [1].”
Reinforcement Learning: An Introduction
Johnson, J. D., Li, J., & Chen, Z. (2000) · Neurocomputing · doi:10.1016/s0925-2312(00)00324-6 · Cited by 8,414
Introduction to Reinforcement Learning
Sutton, R. S., & Barto, A. G. (1998) · The MIT Press eBooks · Cited by 6,855
Internet of Things for Smart Cities
Zanella, A., Bui, N., Castellani, A., Vangelista, L., & Zorzi, M. (2014) · IEEE Internet of Things Journal · doi:10.1109/jiot.2014.2306328 · Cited by 6,292 · cited 1×
“With the recent development of the Internet of Things (IoT) [5, 11, 43] and the Internet of Vehicles (IoV) [22, 41], the fully distributed execution could be practically deployed by distributing the centralized trained policy to each vehicle through the Vehicle-to-Network (V2N) [1].”
An Introduction to MultiAgent Systems
Wooldridge, M. (2002) · Cited by 5,394
Learning to predict by the methods of temporal differences
Sutton, R. S. (1988) · Machine Learning · doi:10.1007/bf00115009 · Cited by 5,148
(1999) · Cited by 5,147
Deterministic Policy Gradient Algorithms
Silver, D., Lever, G., Heess, N., Degris, T., Wierstra, D., & Riedmiller, M. (2014) · HAL · Cited by 4,896
Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments
Lowe, R., Wu, Y., Tamar, A., Harb, J., Abbeel, P., & Mordatch, I. (2017) · arXiv · doi:10.48550/arxiv.1706.02275 · Cited by 3,377
Markov games as a framework for multi-agent reinforcement learning
Littman, M. L. (1994) · Machine Learning Proceedings 1994 · doi:10.1016/b978-1-55860-335-6.50027-1 · Cited by 2,614 · cited 1×
“cooperative order dispatching algorithm (COD). 2.1 Order Dispatching as a Markov Game 2.1.1 Game Settings.We model the order dispatching task by a Partially Observable Markov Decision Process (POMDP) [20] in a fullycooperativesetting,definedbyatupleΓ= ⟨S,P,A,R,O,N,γ⟩, where S,P,A,R,O,N,γare the sets of states, transit…”
MethodInfluentialCounterfactual Multi-Agent Policy Gradients
Foerster, J., Farquhar, G., Afouras, T., Nardelli, N., & Whiteson, S. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11794 · Cited by 1,873
Multi-Agent Reinforcement Learning: Independent vs. Cooperative Agents
Tan, M. (1993) · Machine Learning Proceedings 1993 · doi:10.1016/b978-1-55860-307-3.50049-6 · Cited by 1,810
An Introduction to Multi-Agent Systems
Balaji, P. G., & Srinivasan, D. (2010) · Studies in Computational Intelligence · doi:10.1007/978-3-642-14435-6_1 · Cited by 1,376
Connected Vehicles: Solutions and Challenges
Lu, N., Cheng, N., Zhang, N., Shen, X., & Mark, J. W. (2014) · IEEE Internet of Things Journal · doi:10.1109/jiot.2014.2327587 · Cited by 1,252 · cited 1×
“ing the order dispatching system more robust to potential hardware or connectivity failures. With the recent development of the Internet of Things (IoT) [5, 11, 43] and the Internet of Vehicles (IoV) [22, 41], the fully distributed execution could be practically deployed by distributing the centralized trained policy…”
BackgroundRidesharing: The state-of-the-art and future directions
Furuhata, M., Dessouky, M., Ordóñez, F., Brunet, M.-E., Wang, X., & Koenig, S. (2013) · Transportation Research Part B: Methodological · doi:10.1016/j.trb.2013.08.012 · Cited by 885 · cited 1×
“Real-time ridesharing refers to the task of helping to arrange onetime shared rides on very short notice [4, 10].”
BackgroundCooperative Multi-agent Control Using Deep Reinforcement Learning
Gupta, J. K., Egorov, M., & Kochenderfer, M. (2017) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-71682-4_5 · Cited by 874 · cited 1×
“d to fail in practice [24] because of the non-stationary nature of the multi-agent environment. Several approaches have been attempted to address this problem, including sharing the policy parameters [12], training the Q-function with other agent’s policy parameters [37], or using importance sampling to learn from dat…”
BackgroundInterworking of DSRC and Cellular Network Technologies for V2X Communications: A Survey
Abboud, K., Omar, H. A., & Zhuang, W. (2016) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2016.2591558 · Cited by 785 · cited 3×
“ess of our reward function setting is empirically verified in Section 3.1.3. •γ: Each agent i aims to maximize its total discounted reward Gt i = Õ ∞ k=t γk−trk i (4) from time step t onwards, whereγ∈[0,1]is the discount factor. We denote joint quantities over agents in bold, and joint quantities over agents other tha…”
BackgroundInfluentialOn the Convergence of Stochastic Iterative Dynamic Programming Algorithms
Jaakkola, T., Jordan, M. I., & Singh, S. P. (1994) · Neural Computation · doi:10.1162/neco.1994.6.6.1185 · Cited by 770
Multi-agent Reinforcement Learning: An Overview
Buşoniu, L., Babuška, R., & De Schutter, B. (2010) · Studies in Computational Intelligence · doi:10.1007/978-3-642-14435-6_7 · Cited by 741 · cited 1×
“that is generally inherent to the centralized architecture. Staying different from these approaches, in this work we model the order dispatching problem with multi-agent reinforcement learning (MARL) [7], where agents share a centralized judge (the critic) to rate their decisions (actions) and update their strategies…”
MethodAn overview of Internet of Vehicles
Yang, F., Wang, S., Li, J., Liu, Z., & Sun, Q. (2014) · China Communications · doi:10.1109/cc.2014.6969789 · Cited by 665 · cited 1×
“ing the order dispatching system more robust to potential hardware or connectivity failures. With the recent development of the Internet of Things (IoT) [5, 11, 43] and the Internet of Vehicles (IoV) [22, 41], the fully distributed execution could be practically deployed by distributing the centralized trained policy…”
BackgroundLarge-Scale Order Dispatch in On-Demand Ride-Hailing Platforms: A Learning and Planning Approach
Xu, Z., Li, Z., Guan, Q., Zhang, D., Li, Q., Nan, J., Liu, C., Bian, W., & Ye, J. (2018) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3219819.3219824 · Cited by 556
Self-citationMatignon, L., Laurent, G. J., & Le Fort-Piat, N. (2012) · The Knowledge Engineering Review · doi:10.1017/s0269888912000057 · Cited by 415 · cited 1×
“They, however, tend to fail in practice [24] because of the non-stationary nature of the multi-agent environment.”
BackgroundStabilising Experience Replay for Deep Multi-Agent Reinforcement Learning
Foerster, J., Nardelli, N., Farquhar, G., Afouras, T., Torr, P. H. S., Kohli, P., & Whiteson, S. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1702.08887 · Cited by 397 · cited 1×
“Several approaches have been attempted to address this problem, including sharing the policy parameters [12], training the Q-function with other agent’s policy parameters [37], or using importance sampling to learn from data gathered in a different environment [9].”
Mean Field Multi-Agent Reinforcement Learning
Yang, Y., Luo, R., Li, M., Zhou, M. Q., Zhang, W., & Wang, J. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.05438 · Cited by 286 · cited 5×
“agent. We can easily know that necessity condition of the equilibrium is φ∗= A−1 ϕ∗ bϕ∗. The existence of the such equilibrium has been restricted in the scenario that meets Assumption 3. Yang et al. [42] proved that under the Assumption 3, the existing equilibrium, either in the form of a global equilibrium or in the…”
BackgroundMethodResultInfluentialSelf-citationPeng, P., Wen, Y., Yang, Y., Yuan, Q., Tang, Z., Long, H., & Wang, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1703.10069 · Cited by 241 · cited 1×
“The idea of centralized training with decentralized execution has been investigated by several works [8, 21, 29] recently for MARL using policy gradients [34], and deep neural networks function approximators, based on the actor-critic framework [15].”
MethodSelf-citationA Taxi Order Dispatch Model based On Combinatorial Optimization
Zhang, L., Hu, T., Min, Y., Wu, G., Zhang, J., Feng, P., Gong, P., & Ye, J. (2017) · Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/3097983.3098138 · Cited by 228 · cited 2×
“[44] proposed a novel model based on centralized combinatorial optimization by concurrently matching multiple driver-order pairs within a short time window.”
BackgroundSelf-citationA Collaborative Multiagent Taxi-Dispatch System
Seow, K. T., Dang, N. H., & Lee, D.-H. (2010) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2009.2028577 · Cited by 193 · cited 2×
“ort-term interests into account. The above methods work with centralized control, which is prone to the potential "single point of failure" [23]. With the decentralized setting, Seow et al. [30] addressed the problem by grouping neighboring drivers and orders in a small multi-agent environment, and then simu…”
BackgroundExtending Q-Learning to General Adaptive Multi-Agent Systems
Tesauro, G. J. (2003) · Cited by 187
Taxi Dispatch System Based on Current Demands and Real-Time Traffic Conditions
Lee, D.-H., Wang, H., Cheu, R. L., & Teo, S. H. (2004) · Transportation Research Record: Journal of the Transportation Research Board · doi:10.3141/1882-23 · Cited by 182 · cited 1×
“[16] and Lee et al.”
A Unified Analysis of Value-Function-Based Reinforcement-Learning Algorithms
Szepesvári, C., & Littman, M. L. (1999) · Neural Computation · doi:10.1162/089976699300016070 · Cited by 179
Multiagent Bidirectionally-Coordinated Nets for Learning to Play StarCraft Combat Games.Preprint
Peng, P., Yuan, Q., Wen, Y., Yang, Y., Tang, Z., Long, H., & Wang, Y.-X. (2017) · arXiv (Cornell University) · Cited by 176
Self-citationDeep Reinforcement Learning with Knowledge Transfer for Online Rides Order Dispatching
Wang, Z., Qin, Z. T., Tang, X., Ye, J., & Hongtu Zhu, B. (2018) · Proceedings · doi:10.1109/icdm.2018.00077 · Cited by 157 · cited 1×
“[38] proposed an RL-based dispatching algorithm to optimize resource utilization and user experience in a global and more farsighted view.”
BackgroundSelf-citationAnalyzing and visualizing multiagent rewards in dynamic and stochastic domains
Agogino, A., & Tumer, K. (2008) · Autonomous Agents and Multi-Agent Systems · doi:10.1007/s10458-008-9046-9 · Cited by 128
Amey, A., Attanucci, J., & Mishalani, R. (2011) · Transportation Research Record: Journal of the Transportation Research Board · doi:10.3141/2217-13 · Cited by 114
An Empirical Analysis of On-demand Ride-sharing and Traffic Congestion
Li, Z., Hong, Y., & Zhang, Z. (2017) · Proceedings of the Annual Hawaii International Conference on System Sciences · doi:10.24251/hicss.2017.002 · Cited by 107 · cited 1×
“By exploiting the data of individual trajectories in both space and time dimensions, it offers more efficiency on traffic management, and the traffic congestion can be further alleviated as well [18].”
BackgroundKonda, V. R., & Tsitsiklis, J. N. (1999) · neural information processing systems · Cited by 45
Cited by
A Survey of Traffic Prediction: from Spatio-Temporal Data to Intelligent Transportation
Yuan, H., & Li, G. (2021) · Data Science and Engineering · doi:10.1007/s41019-020-00151-z · Cited by 389
Coordinating matching, rebalancing and charging of electric ride-hailing fleet under hybrid requests
Yu, X., Zhu, Z., Mao, H., Hua, M., Li, D., Chen, J., & Xu, H. (2023) · Transportation Research Part D: Transport and Environment · doi:10.1016/j.trd.2023.103903 · Cited by 12
Metadata
- Authors
- Li, M., Qin, Z., Jiao, Y., Yang, Y., Wang, J., Wang, C., Wu, G., & Ye, J.
- Venue
- The World Wide Web Conference
- Publisher
- ACM
- Published
- 2019-05-13
- Pages
- 983-994
- OpenAlex
- W2914154006
- Cited by
- 237
- References
- 39
- FWCI
- 32.65
- Type
- conference paper
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.