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...
详细信息
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.
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. ...
详细信息
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.
Geometric self-reconfiguration sequences and a distributed control method are developed for the modular robot M-TRAN. Large scale self-reconfiguration contains several clusters of modules moving in parallel. A decentr...
详细信息
ISBN:
(纸本)078039044X
Geometric self-reconfiguration sequences and a distributed control method are developed for the modular robot M-TRAN. Large scale self-reconfiguration contains several clusters of modules moving in parallel. A decentralized and asynchronous control is suitable for such a task. A multi thread type simulation program was developed to design and verify such self-reconfigu ration sequences. A distributed controller system was developed for the new M-TRAN hardware, which can realize decentralized and asynchronous control of modules. Its basic functions were verified by experiments.
We present a novel type of modular robotic system that has a capability of both autonomous shape reconfiguration and robotic motion generation. We have developed hardware modules and examined basic mechanical function...
详细信息
ISBN:
(纸本)0819438618
We present a novel type of modular robotic system that has a capability of both autonomous shape reconfiguration and robotic motion generation. We have developed hardware modules and examined basic mechanical functions and reconfiguration. A simulation system of the modular robotic system has also been developed to design the reconfiguration sequence and motion for a cluster of the modules.
Growing complexity of artificial systems arises reliability and flexibility issues of large system design. Robots are not exception of this, and many attempts have been made to realize reliable and flexible robot syst...
详细信息
Growing complexity of artificial systems arises reliability and flexibility issues of large system design. Robots are not exception of this, and many attempts have been made to realize reliable and flexible robot systems. distributed modular composition of robot is one of the most effective approaches to attain such abilities and has a potential to adapt to its surroundings by changing its configuration autonomously according to information of surroundings. In this paper, we propose a novel three-dimensional self-reconfigurable robotic module. Each module has a very simple structure that consists of two semi-cylindrical parts connected by a link. The modular system is capable of not only building static structure but also generating dynamic robotic motion. We present details of the mechanical/electrical design of the developed module and its control system architecture. Experiments using ten modules with centralized control demonstrate robotic configuration change, crawling locomotion and three types of quadruped locomotion.
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 ...
详细信息
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.
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 syst...
详细信息
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.
A three-dimensional (3D) self-reconfigurable system made of identical units is proposed. Each unit has six arms on the surface of its base cube which can connect to neighboring units mechanically. By the connection, a...
详细信息
暂无评论