Fuzzy policy reinforcement learning in cooperative multi-robot systems

Dongbing Gu, Erfu Yang

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

A multi-agent reinforcement learning algorithm with fuzzy policy is addressed in this paper. This algorithm is used to deal with some control problems in cooperative multi-robot systems. Specifically, a leader-follower robotic system and a flocking system are investigated. In the leader-follower robotic system, the leader robot tries to track a desired trajectory, while the follower robot tries to follow the reader to keep a formation. Two different fuzzy policies are developed for the leader and follower, respectively. In the flocking system, multiple robots adopt the same fuzzy policy to flock. Initial fuzzy policies are manually crafted for these cooperative behaviors. The proposed learning algorithm finely tunes the parameters of the fuzzy policies through the policy gradient approach to improve control performance. Our simulation results demonstrate that the control performance can be improved after the learning.

LanguageEnglish
Pages7-22
Number of pages16
JournalJournal of Intelligent and Robotic Systems
Volume48
Issue number1
DOIs
Publication statusPublished - 1 Jan 2007

Fingerprint

Reinforcement learning
Robots
Learning algorithms
Robotics
Trajectories

Keywords

  • cooperative control
  • flocking behavior
  • multi-agent reinforcement learning
  • policy gradient reinforcement learning

Cite this

@article{2e585d9c82c44e4f8c67e3665a05c798,
title = "Fuzzy policy reinforcement learning in cooperative multi-robot systems",
abstract = "A multi-agent reinforcement learning algorithm with fuzzy policy is addressed in this paper. This algorithm is used to deal with some control problems in cooperative multi-robot systems. Specifically, a leader-follower robotic system and a flocking system are investigated. In the leader-follower robotic system, the leader robot tries to track a desired trajectory, while the follower robot tries to follow the reader to keep a formation. Two different fuzzy policies are developed for the leader and follower, respectively. In the flocking system, multiple robots adopt the same fuzzy policy to flock. Initial fuzzy policies are manually crafted for these cooperative behaviors. The proposed learning algorithm finely tunes the parameters of the fuzzy policies through the policy gradient approach to improve control performance. Our simulation results demonstrate that the control performance can be improved after the learning.",
keywords = "cooperative control, flocking behavior, multi-agent reinforcement learning, policy gradient reinforcement learning",
author = "Dongbing Gu and Erfu Yang",
year = "2007",
month = "1",
day = "1",
doi = "10.1007/s10846-006-9103-z",
language = "English",
volume = "48",
pages = "7--22",
journal = "Journal of Intelligent and Robotic Systems",
issn = "0921-0296",
number = "1",

}

Fuzzy policy reinforcement learning in cooperative multi-robot systems. / Gu, Dongbing; Yang, Erfu.

In: Journal of Intelligent and Robotic Systems, Vol. 48, No. 1, 01.01.2007, p. 7-22.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fuzzy policy reinforcement learning in cooperative multi-robot systems

AU - Gu, Dongbing

AU - Yang, Erfu

PY - 2007/1/1

Y1 - 2007/1/1

N2 - A multi-agent reinforcement learning algorithm with fuzzy policy is addressed in this paper. This algorithm is used to deal with some control problems in cooperative multi-robot systems. Specifically, a leader-follower robotic system and a flocking system are investigated. In the leader-follower robotic system, the leader robot tries to track a desired trajectory, while the follower robot tries to follow the reader to keep a formation. Two different fuzzy policies are developed for the leader and follower, respectively. In the flocking system, multiple robots adopt the same fuzzy policy to flock. Initial fuzzy policies are manually crafted for these cooperative behaviors. The proposed learning algorithm finely tunes the parameters of the fuzzy policies through the policy gradient approach to improve control performance. Our simulation results demonstrate that the control performance can be improved after the learning.

AB - A multi-agent reinforcement learning algorithm with fuzzy policy is addressed in this paper. This algorithm is used to deal with some control problems in cooperative multi-robot systems. Specifically, a leader-follower robotic system and a flocking system are investigated. In the leader-follower robotic system, the leader robot tries to track a desired trajectory, while the follower robot tries to follow the reader to keep a formation. Two different fuzzy policies are developed for the leader and follower, respectively. In the flocking system, multiple robots adopt the same fuzzy policy to flock. Initial fuzzy policies are manually crafted for these cooperative behaviors. The proposed learning algorithm finely tunes the parameters of the fuzzy policies through the policy gradient approach to improve control performance. Our simulation results demonstrate that the control performance can be improved after the learning.

KW - cooperative control

KW - flocking behavior

KW - multi-agent reinforcement learning

KW - policy gradient reinforcement learning

UR - http://www.scopus.com/inward/record.url?scp=33846038724&partnerID=8YFLogxK

U2 - 10.1007/s10846-006-9103-z

DO - 10.1007/s10846-006-9103-z

M3 - Article

VL - 48

SP - 7

EP - 22

JO - Journal of Intelligent and Robotic Systems

T2 - Journal of Intelligent and Robotic Systems

JF - Journal of Intelligent and Robotic Systems

SN - 0921-0296

IS - 1

ER -