Название: Multi-Agent Coordination: A Reinforcement Learning ApproachАвтор: Arup Kumar Sadhu, Amit Konar
Издательство: Wiley-IEEE Press
Год: 2021
Страниц: 315
Язык: английский
Формат: pdf (true)
Размер: 12.5 MB
Discover the latest developments in multi-robot coordination techniques with this insightful and original resource.
Multi-Agent Coordination: A Reinforcement Learning Approach delivers a comprehensive, insightful, and unique treatment of the development of multi-robot coordination algorithms with minimal computational burden and reduced storage requirements when compared to traditional algorithms. The accomplished academics, engineers, and authors provide readers with both a high-level introduction to, and overview of, multi-robot coordination, and in-depth analyses of learning-based planning algorithms.
You'll learn about how to accelerate the exploration of the team-goal and alternative approaches to speeding up the convergence of TMAQL by identifying the preferred joint action for the team. The authors also propose novel approaches to consensus Q-learning that address the equilibrium selection problem and a new way of evaluating the threshold value for uniting empires without imposing any significant computation overhead. Finally, the book concludes with an examination of the likely direction of future research in this rapidly developing field.
In recent times, researchers are taking keen interest to employ machine learning in multi-agent cooperation. The primary advantage of machine learning is to generate the action plans in sequence from the available sensory readings of the robots. In case of a single robot, learning the action plans from the sensory readings is straightforward.
Because of the difficulty of generating training instances and excessive computational overhead to learn those instances, coupled with the need for handling dynamic situations, researchers felt the importance of reinforcement learning (RL). In RL, we need not provide any training instance, but employ a critic who provides a feedback to the learning algorithm about the possible reward/penalty of the actions by the agent. The agent/s on receiving the approximate measure of penalty/reward understands which particular sensory-motor instances they need to learn for future planning applications. The dynamic nature of environment thus can easily be learned by RL. In the multi-agent scenario, RL needs to take care of learning in joint state/action space of the agents. Here, each agent learns the sensory-motor instances in the joint state/action space with an ultimate motive to learn the best actions for itself to optimize its rewards.
Readers will discover cutting-edge techniques for multi-agent coordination, including:
An introduction to multi-agent coordination by reinforcement learning and evolutionary algorithms, including topics like the Nash equilibrium and correlated equilibrium
Improving convergence speed of multi-agent Q-learning for cooperative task planning
Consensus Q-learning for multi-agent cooperative planning
The efficient computing of correlated equilibrium for cooperative q-learning based multi-agent planning
A modified imperialist competitive algorithm for multi-agent stick-carrying applications
Perfect for academics, engineers, and professionals who regularly work with multi-agent learning algorithms, Multi-Agent Coordination: A Reinforcement Learning Approach also belongs on the bookshelves of anyone with an advanced interest in machine learning and artificial intelligence as it applies to the field of cooperative or competitive robotics.
Скачать Multi-Agent Coordination: A Reinforcement Learning Approach