# Multiagent reinforcement learning algorithms for multiple-UAV confrontation
This is the source code of "Efficient training techniques for multi-agent reinforcement learning in combatant tasks",
we construct a multi-agent confrontation environment originated from a combatant scenario of multiple unman aerial vehicles.
To begin with, we consider to solve this confrontation problem with two types of MARL algorithms.
One is extended from the classical deep Q-network for multi-agent settings (MADQN).
The other one is extended from the state-of-art multi-agent reinforcement method, multi-agent deep deterministic policy gradient (MADDPG).
We compare the two methods for the initial confrontation scenario and find that MADDPG outperforms MADQN.
Then with MADDPG as the baseline, we propose three efficient training techniques, i.e., scenario-transfer training, self-play training and rule-coupled training.
![image](https://github.com/sanjinzhi/multiagent-confrontation/blob/master/Rule-coupled%20vs%20Random.gif)
Rule-coupled red agents vs Random-move blue agents
![image](https://github.com/sanjinzhi/multiagent-confrontation/blob/master/Rule-coupled%20vs%20Selfplay.gif)
Rule-coupled red agents vs Blue agents trained by self-play
没有合适的资源?快使用搜索试试~ 我知道了~
强化学习无人机对抗附python代码.zip
共111个文件
py:54个
pyc:16个
txt:13个
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 163 浏览量
2024-05-21
23:30:33
上传
评论
收藏 5.31MB ZIP 举报
温馨提示
3.代码特点:参数化编程、参数可方便更改、代码编程思路清晰、注释明细。 4.适用对象:计算机,电子信息工程、数学等专业的大学生课程设计、期末大作业和毕业设计。
资源推荐
资源详情
资源评论
收起资源包目录
强化学习无人机对抗附python代码.zip (111个子文件)
checkpoint 167B
.DS_Store 6KB
.DS_Store 6KB
.DS_Store 6KB
Rule-coupled vs Random.gif 3.29MB
Rule-coupled vs Selfplay.gif 2.22MB
.gitignore 84B
tag-dqn_21500_1.h5 27KB
tag-dqn_21500_0.h5 27KB
tag-dqn_21500_2.h5 27KB
openai-maddpg.iml 398B
run_parameters.json 368B
settings.json 76B
settings.json 76B
README.md 1KB
README.md 745B
README.md 744B
README.md 2B
not-zip-safe 1B
not-zip-safe 1B
PKG-INFO 280B
PKG-INFO 262B
environment.py 16KB
competition_3v3.py 16KB
environment-tmp.py 15KB
competition_3v3-tmp.py 13KB
environment.py 13KB
tf_util.py 12KB
distributions.py 12KB
simple_world_comm.py 11KB
simple_world_comm.py 11KB
rendering.py 11KB
rendering.py 11KB
angle_3v3.py 10KB
maddpg.py 9KB
dqn_tag.py 7KB
core.py 7KB
core.py 7KB
simple_crypto.py 6KB
simple_crypto.py 6KB
simple_tag_v1.py 6KB
simple_adversary.py 6KB
simple_adversary.py 6KB
simple_tag_v1.py 6KB
simple_tag.py 6KB
simple_push.py 4KB
simple_push.py 4KB
simple_spread.py 4KB
simple_spread.py 4KB
simple_speaker_listener.py 3KB
simple_speaker_listener.py 3KB
simple_reference.py 3KB
simple_reference.py 3KB
replay_buffer.py 3KB
dqn.py 2KB
multi_discrete.py 2KB
multi_discrete.py 2KB
simple.py 2KB
simple.py 2KB
make_env.py 2KB
policy.py 2KB
policy.py 2KB
interactive.py 1KB
interactive.py 1KB
__init__.py 467B
__init__.py 467B
setup.py 426B
setup.py 426B
setup.py 408B
__init__.py 407B
scenario.py 309B
scenario.py 309B
__init__.py 145B
__init__.py 145B
__init__.py 0B
__init__.py 0B
distributions.cpython-35.pyc 21KB
rendering.cpython-35.pyc 16KB
tf_util.cpython-35.pyc 13KB
environment.cpython-35.pyc 11KB
competition_3v3.cpython-35.pyc 10KB
maddpg.cpython-35.pyc 7KB
core.cpython-35.pyc 6KB
simple_tag_yuan_v2.cpython-35.pyc 6KB
simple_tag_v1.cpython-35.pyc 5KB
replay_buffer.cpython-35.pyc 4KB
multi_discrete.cpython-35.pyc 3KB
simple.cpython-35.pyc 2KB
__init__.cpython-35.pyc 1KB
scenario.cpython-35.pyc 668B
__init__.cpython-35.pyc 459B
__init__.cpython-35.pyc 434B
reward setting 219B
LICENSE.txt 1KB
LICENSE.txt 1KB
LICENSE.txt 1KB
SOURCES.txt 880B
SOURCES.txt 215B
新建.txt 99B
readme.txt 34B
共 111 条
- 1
- 2
资源评论
matlab科研助手
- 粉丝: 2w+
- 资源: 2696
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功