检索结果-内蒙古大学图书馆

Bottom-up multi-agent reinforcement learning by reward shaping for cooperative-competitive tasks

APPLIED INTELLIGENCE 2021年第7期51卷 4434-4452页

作者： Aotani, Takumi Kobayashi, Taisuke Sugimoto, Kenji Nara Inst Sci & Technol Div Informat Sci 8916-5 Takayama Cho Nara Japan

A multi-agent system (MAS) is expected to be applied to various real-world problems where a single agent cannot accomplish given tasks. Due to the inherent complexity in the real-world MAS, however, manual design of group behaviors of agents is intractable. Multi-agent reinforcement learning (MARL), which is a framework for multiple agents in the same environment to learn their policies adaptively by using reinforcement learning, would be a promising methodology for such complexity in the MAS. To acquire the group behaviors by MARL, all the agents are required to understand how to achieve the respective tasks cooperatively. So far, we have proposed "bottom-up MARL", which is a decentralized system to manage real and large-scale MARL, with a reward shaping algorithm to represent the group behaviors. The reward shaping algorithm, however, assumes that all the agents are in cooperative relationships to some extent. In this paper, therefore, we extend this algorithm to allow the agents not to know the interests between them. The interests are regarded as correlation coefficients derived from the agents' rewards, which are numerically estimated in an online manner. Actually, in both simulations and real experiments without knowledge of the interests between the agents, they correctly estimated their interests, thereby allowing them to derive their new rewards to represent the feasible group behaviors in the decentralized manner. As a result, our extended algorithm succeeded in acquiring the group behaviors from cooperative tasks to competitive tasks.

关键词： distributed autonomous system Reinforcement learning Reward shaping Interests between agents

来源：评论

学校读者我要写书评

暂无评论

distributed autonomous Control system of the Smart Grid Based on the Turing Pattern Model

引用

IEEJ TRANSACTIONS ON ELECTRICAL AND ELECTRONIC ENGINEERING 2013年第6期8卷 649-651页

作者： Ishida, Takeshi Natl Fisheries Univ Shimonoseki Yamaguchi 7596595 Japan

Renewable energy resources, which are generally unstable, affect the performance of electricity grids when they are fully connected. A distributed energy supply network system is expected to compensate for the anticipated performance problems. We have constructed a distributed autonomous system that controls batteries installed in each house. In this system, even if central control does not exist, electrical charging and discharging of the batteries are carried out for self-organization. As a result, the grid stability is maintained in the local urban district. Our system is based on the Turing pattern model. (c) 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

关键词： smart grid photovoltaic cell secondary battery distributed autonomous system Turing pattern model

来源：评论

学校读者我要写书评

暂无评论

MAT-DQN: Toward Interpretable Multi-agent Deep Reinforcement Learning for Coordinated Activities 30th

MAT-DQN: Toward Interpretable Multi-agent Deep Reinforcement...

引用

30th International Conference on Artificial Neural Networks (ICANN)

作者： Motokawa, Yoshinari Sugawara, Toshiharu Waseda Univ Dept Comp Sci & Engn Tokyo Japan

ISBN: (纸本)9783030863807;9783030863791

We propose an interpretable neural network architecture for multi-agent deep reinforcement learning to understand the rationale for learned cooperative behavior of the agents. Although the deep learning technology has contributed significantly to multi-agent systems to build coordination among agents, it is still unclear what information the agents depend on to behave cooperatively. Removing this ambiguity may further improve the efficiency and productivity of multi-agent systems. The main idea of our proposal is to adopt the transformer to deep Q-network for addressing the above-mentioned issue. By extracting multi-head attention weights from the transformer encoder, we propose a multi-agent transformer deep Q-network (MAT-DQN) and show that agents using attention mechanisms possess better coordination capability with other agents despite being trained individually for a cooperative patrolling task problem;thus, they can exhibit better performance results compared with the agents with vanilla DQN (which is a baseline method). Furthermore, we indicate that it is possible to visualize heatmaps of attentions, which indicate the influential input-information in agents' decision-making process for their cooperative behaviors.

关键词： distributed autonomous system Multi-agent deep reinforcement learning Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Bottom-up Multi-agent Reinforcement Learning for Selective Cooperation

Bottom-up Multi-agent Reinforcement Learning for Selective C...

引用

IEEE International Conference on systems, Man, and Cybernetics (SMC)

作者： Aotani, Takumi Kobayashi, Taisuke Sugimoto, Kenji Nara Inst Sci & Technol Grad Sch Informat Sci Nara Japan Nara Inst Sci & Technol Grad Sch Sci & Technol Div Informat Sci Nara Japan

ISBN: (纸本)9781538666500

Applications of multi-agent system like cooperative transport are found in various domains of real world. Due to the complexity inherent in multi-agent system, however, handling with preprogramming is difficult. Multi-agent reinforcement learning (MARL), which is a framework to make multiple agents in the same environment learn their policies simultaneously using reinforcement learning, is receiving attention. In the conventional MARL, although decentralization is essential for feasible learning, rewards for the agents have been allocated from a centralized system in the environment. Instead of such "top-down" MARL, to achieve the completely distributed autonomous systems, we tackle a new paradigm named "bottom-up" MARL, where the agents get their own rewards. The bottom-up MARL requires to share the respective rewards for emerging orderly group behaviors, which cannot be acquired merely by maximizing the mean of them. We therefore propose the architecture that has three components: estimating rewards of other agents;selecting rewards to reinforce from the correlation, and;promoting the exploration to find unknown correlation. The proposed architecture is verified that every element is essential by numerical simulation performed in stages. A similar task is also accomplished in dynamical simulation under the same conditions as the actual robots.

关键词： distributed autonomous system Reinforcement leaning Stochastic neural network

来源：评论

学校读者我要写书评

暂无评论

A distributed autonomous Approach to Developing a Disaster Evacuation Assist system

引用

JOURNAL OF DISASTER RESEARCH 2015年第6期10卷 1081-1090页

作者： Iizuka, Yasuki Kinoshita, Katsuya Iizuka, Kayo Tokai Univ Dept Math Sci 4-1-1 Kitakaname Hiratsuka Kanagawa 2591292 Japan Syst Res Co LTD Nakamura Ku Nagoya Aichi 4530861 Japan Senshu Univ Sch Network & Informat Tama Ku Kawasaki Kanagawa 2148580 Japan

In times of disaster, or other emergency situations, it is essential for people to be evacuated in a smooth manner. Evacuation must be performed promptly and safely. It is necessary to avoid generating a secondary disaster at the time of evacuation. However, this is not easy to realize, because people often tend to panic when faced with disaster, crowding the evacuation passageways of buildings. On the other hand, people do not attempt to evacuate themselves from danger when the normalcy bias has occurred. Therefore, evacuation guidance is very important. However, it is impossible to guide all evacuees through authorities such as disaster countermeasure offices. To deal with this issue, the authors propose a system that provides optimal evacuation guidance autonomously without central server. The system works on the mobile devices of evacuees, performs distributed calculations using the framework of the distributed constraint optimization problem on ad-hoc communication, and does not need a central server. In the experiment using multi-agent simulation, for the case where the evacuees can receive evacuation guidance from this system, the evacuation completion time decreased. This paper presents an overview and the evaluation results of the prototype of the disaster evacuation assistance system.

关键词： disaster evacuation distributed autonomous system DCOP multi-agent simulation

来源：评论

学校读者我要写书评

暂无评论

Disaster Information Collecting/Providing Service for Local Residents

Disaster Information Collecting/Providing Service for Local ...

引用

Symposium on Human Interface/14th International Conference on Human-Computer Interaction (HCI)

作者： Takahashi, Yuichi Kobayashi, Daiji Yamamoto, Sakae Tokyo Univ Sci Dept Management Sci 1-3 Kagurazaka Shinjuku Tokyo 162 Japan Chitose Inst Sci & Technol Fac Photon Sci Dept Dept Global Syst Design Chitose Hokkaido Japan

ISBN: (纸本)9783642216688;9783642216695

It has been pointed out that when people lack the information needed in the event of a disaster, such as a disastrous earthquake, this could lead to social chaos, including unwanted rumors and outrages, or could disrupt rescue and relief activities(1, 2). In Japan, by law in principle, self-help or mutual assistance is required immediately after a disaster, and local residents are required to make judgments for action on their own. Although disaster information systems are gradually being organized at the municipal level, actual emergency evacuation areas and essential information for local citizens are still not sufficiently ready for provision at this stage.(3) In this study, we established and evaluated a service infrastructure with an autonomous wireless network, aiming at providing services to collect and deliver disaster information, which will be required by local residents.

关键词： earthquake disaster victims distributed autonomous system wireless network

来源：评论

学校读者我要写书评

暂无评论

A Development of Information system for Disaster Victims with autonomous Wireless Network

A Development of Information System for Disaster Victims wit...

引用

Symposium on Human Interface held at the HCI International 2009

作者： Takahashi, Yuichi Kobayashi, Daiji Yamamoto, Sakae Tokyo Univ Sci Dept Management Sci Shinjuku Ku 1-3 Kagurazaka Tokyo 162 Japan Chitose Inst Sci & Technol Fac Photon Sci Dept Global Syst Design Chitose Hokkaido 75865 Japan

ISBN: (纸本)9783642025587

In times of huge disaster such as earthquake, information needs increase among victims and rescuers. Social ferment rises within afflicted area and the damage is spread, if the needs of information are not satisfied. In this research we developed an information system for disaster victims as a distributed autonomous system using it wireless network. This system consists of many sub systems. These Sub systems are robust for collecting disaster information because they are small and simple. An authorized user can register information using one of the sub systems that is working correctly. Asynchronously, they search another sub system via wireless network, and then they communicate to each other in order to exchange information they have. As a result, the information will be shared within a wide area by those processes like a bucket brigade.

关键词： earthquake disaster victims distributed autonomous system wireless network

来源：评论

学校读者我要写书评

暂无评论

distributed self-reconfiguration of M-TRAN III modular robotic system

引用

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH 2008年第3-4期27卷 373-386页

作者： Kurokawa, Haruhisa Tomita, Kohji Kamimura, Akiya Kokaji, Shigeru Hasuo, Takashi Murata, Satoshi Natl Inst Adv Ind Sci & Technol Intelligent Syst Res Inst Tsukuba Ibaraki 3058564 Japan Univ Tsukuba Grad Sch Syst & Informat Engn Tsukuba Ibaraki 3058573 Japan Tokyo Inst Technol Interdisciplinary Grad Sch Sci & Engn Yokohama Kanagawa 2268502 Japan

A new prototype of a self-reconfigurable modular robot, M-TRAN III, has been developed, with an improved fast and rigid connection mechanism. Using a distributed controller, various control modes are possible: single-master, globally synchronous control or parallel asynchronous control. Self-reconfiguration experiments using up to 24 modules were undertaken by centralized or decentralized control. Experiments using decentralized control examined a modular structure moved in a given direction as a flow produced by local self-reconfigurations. In all experiments, system homogeneity and scalability were maintained: modules used identical software except for their ID numbers. Identical self-reconfiguration was realized when different modules were used in initial configurations.

关键词： distributed autonomous system decentralized control controller area network

来源：评论

学校读者我要写书评

暂无评论

Analysis of coupled van der Pol oscillators and implementation to a myriapod robot

引用

IFAC Proceedings Volumes 2008年第2期41卷 767-772页

作者： Naoki KUWATA Yoshikatsu HOSHI Ben T. NOHARA Musashi Institute of Technology Tokyo 158-8557 JAPAN

According to the study of nervous system ethology, it is thought that a walk movement of an animal is controlled by Central Pattern Generator (CPG). There are a lot of studies to try the control of the leg type robot based on CPG principle. However, most of the studies consider two or four leg type robots, and there are not many studies performed for robots more than six legs. One of its factor is probably difficulty of constituting stable and periodic CPG in the case of many number of the legs. Therefore, this paper proposes the novel CPG model which is based on coupled van del Pol oscillators. And this paper reports the analysis of proposed coupled van del Pol oscillators model and the implementation result to an actual myriapod robot.

关键词： Coupled van der Pol Oscillators Myriapod Robot Central Pattern Generator Legged Robot distributed autonomous system

来源：评论

学校读者我要写书评

暂无评论

Self-reconfigurable modular robot M-TRAN: distributed control and communication 07

Self-reconfigurable modular robot M-TRAN: distributed contro...

引用

Proceedings of the 1st international conference on Robot communication and coordination

作者： Haruhisa Kurokawa Kohji Tomita Akiya Kamimura Shigeru Kokaji Takashi Hasuo Satoshi Murata National Institute of Advanced Industrial Science and Technology (AIST) Tsukuba Japan University of Tsukuba Tsukuba Japan Tokyo Institute of Technology Yokohama Japan

ISBN: (纸本)9789639799080

M-TRAN is a self-reconfigurable modular robot: each module has an independent battery, two-degree-of-freedom motion, six-surface-connection capability, and intelligence with inter-module communication. The M-TRAN system can perform flexible and adaptive locomotion in various configurations using coordination control based on a central pattern generator (CPG). Various structures of several modules can perform metamorphosis, such as that between a four-legged robot and a snake-like one. In addition to these self-reconfigurations with synchronous control, M-TRAN structures having regularity can move using parallel distributed control and message exchange via the network bus. Self-reconfiguration using infrared local communication has been attempted to improve the system's scalability.

关键词： central pattern generator coordinated motion control distributed autonomous system

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：