Searcharxiv⌕ Search

arXiv subjects

Chunwei Song

Publications and source records attributed to Chunwei Song.

7 recordsLinked to original sources
arXiv2022

Multi-robot Social-aware Cooperative Planning in Pedestrian Environments Using Multi-agent Reinforcement Learning

Safe and efficient co-planning of multiple robots in pedestrian participation environments is promising for applications. In this work, a novel multi-robot social-aware efficient cooperative planner that on the basis of off-policy multi-agent reinforcement learning (MARL) under partial dimension-varying observation and imperfect perception conditions is proposed. We adopt temporal-spatial graph (TSG)-based social encoder to better extract the importance of social relation between each robot and the pedestrians in its field of view (FOV). Also, we introduce K-step lookahead reward setting in multi-robot RL framework to avoid aggressive, intrusive, short-sighted, and unnatural motion decisions generated by robots. Moreover, we improve the traditional centralized critic network with multi-head global attention module to better aggregates local observation information among different robots to guide the process of individual policy update. Finally, multi-group experimental results verify the effectiveness of the proposed cooperative motion planner.

cs.RO↗
arXiv2021

Multi-agent Soft Actor-Critic Based Hybrid Motion Planner for Mobile Robots

In this paper, a novel hybrid multi-robot motion planner that can be applied under non-communication and local observable conditions is presented. The planner is model-free and can realize the end-to-end mapping of multi-robot state and observation information to final smooth and continuous trajectories. The planner is a front-end and back-end separated architecture. The design of the front-end collaborative waypoints searching module is based on the multi-agent soft actor-critic algorithm under the centralized training with decentralized execution diagram. The design of the back-end trajectory optimization module is based on the minimal snap method with safety zone constraints. This module can output the final dynamic-feasible and executable trajectories. Finally, multi-group experimental results verify the effectiveness of the proposed motion planner.

cs.RO↗
arXiv2021

A review of mobile robot motion planning methods: from classical motion planning workflows to reinforcement learning-based architectures

Motion planning is critical to realize the autonomous operation of mobile robots. As the complexity and randomness of robot application scenarios increase, the planning capability of the classical hierarchical motion planners is challenged. With the development of machine learning, deep reinforcement learning (DRL)-based motion planner has gradually become a research hotspot due to its several advantageous features. DRL-based motion planner is model-free and does not rely on the prior structured map. Most importantly, DRL-based motion planner achieves the unification of the global planner and the local planner. In this paper, we provide a systematic review of various motion planning methods. First, we summarize the representative and state-of-the-art works for each submodule of the classical motion planning architecture and analyze their performance features. Subsequently, we concentrate on summarizing RL-based motion planning approaches, including motion planners combined with RL improvements, map-free RL-based motion planners, and multi-robot cooperative planning methods. Last but not least, we analyze the urgent challenges faced by these mainstream RL-based motion planners in detail, review some state-of-the-art works for these issues, and propose suggestions for future research.

cs.RO↗
arXiv2020

Maximizing the number of maximal independent sets of a fixed size

For a fixed graph G, a maximal independent set is an independent set that is not a proper subset of any other independent set. P. Erd\"os, and independently, J. W. Moon and L. Moser, and R. E. Miller and D. E. Muller, determined the maximum number of maximal independent sets in a graph on n vertices, as well as the extremal graphs. In this paper we maximize the number of maximal independent sets of a fixed size for all graphs of order n and determine the extremal graphs. Our result generalizes the classical result.

math.CO↗
arXiv2019

On Combinatorial Rectangles with Minimum $\infty$-Discrepancy

A combinatorial rectangle may be viewed as a matrix whose entries are all +-1. The discrepancy of an m by n matrix is the maximum among the absolute values of its m row sums and n column sums. In this paper, we investigate combinatorial rectangles with minimum discrepancy (0 or 1 for each line depending on the parity). Specifically, we get explicit formula for the number of matrices with minimum L^infinity-discrepancy up to 4 rows, and establish the order of magnitude of the number of such matrices with m rows and n columns while m is fixed and n approaches infinity. By considering the number of column-good matrices with a fixed row-sum vector, we have developed a theory of decreasing criterion on based row-sum vectors with majorization relation, which turns out to be a helpful tool in the proof of our main theorems.

math.CO↗
arXiv2007

A comment on Ryser's conjecture for intersecting hypergraphs

Let $τ(\mathcal{H})$ be the cover number and $ν(\mathcal{H})$ be the matching number of a hypergraph $\mathcal{H}$. Ryser conjectured that every $r$-partite hypergraph $\mathcal{H}$ satisfies the inequality $τ(\mathcal{H}) \leq (r-1) ν(\mathcal{H})$. This conjecture is open for all $r \ge 4$. For intersecting hypergraphs, namely those with $ν(\mathcal{H})=1$, Ryser's conjecture reduces to $τ(\mathcal{H}) \leq r-1$. Even this conjecture is extremely difficult and is open for all $ r \ge 6$. For infinitely many $r$ there are examples of intersecting $r$-partite hypergraphs with $τ(\mathcal{H})=r-1$, demonstrating the tightness of the conjecture for such $r$. However, all previously known constructions are not optimal as they use far too many edges. How sparse can an intersecting $r$-partite hypergraph be, given that its cover number is as large as possible, namely $τ(\mathcal{H}) \ge r-1$? In this paper we solve this question for $r \le 5$, give an almost optimal construction for $r=6$, prove that any $r$-partite intersecting hypergraph with $τ(H) \ge r-1$ must have at least $(3-\frac{1}{\sqrt{18}})r(1-o(1)) \approx 2.764r(1-o(1))$ edges, and conjecture that there exist constructions with $Θ(r)$ edges.

math.CO↗