Multi-Agent Path Finding via Tree LSTM
Not on the map. Why?
Abstract (arXiv)
In recent years, Multi-Agent Path Finding (MAPF) has attracted attention from the fields of both Operations Research (OR) and Reinforcement Learning (RL). However, in the 2021 Flatland3 Challenge, a competition on MAPF, the best RL method scored only 27.9, far less than the best OR method. This paper proposes a new RL solution to Flatland3 Challenge, which scores 125.3, several times higher than the best RL solution before. We creatively apply a novel network architecture, TreeLSTM, to MAPF in our solution. Together with several other RL techniques, including reward shaping, multiple-phase training, and centralized control, our solution is comparable to the top 2-3 OR methods.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 9 of its 15 references that are in the map. First posted on arXiv 24 Oct 2022.
Builds on
9 of 15 references.
References · 9
- 2017Attention Is All You Needcited by 60,361
- 2017Proximal Policy Optimization Algorithmscited by 15,215
- 2019Dota 2 with Large Scale Deep Reinforcement Learningcited by 1,103
- 2015Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networkscited by 306
- 2020Flatland-RL : Multi-Agent Reinforcement Learning on Trainscited by 32
- 2016Multi-Agent Path Finding with Delay Probabilitiescited by 18
- 2021Flatland Competition 2020: MAPF and MARL for Efficient Train Coordination on a Grid Worldcited by 10
- 2018Multi-Agent Path Finding with Deadlinescited by 7
- 2019On the Computational Complexity of Multi-Agent Pathfinding on Directed Graphscited by 7
Built on by
Citing papers per year: 2026: 1.
1 paper.