Abstract:
In this paper, we propose fuzzy logic-based cooperative reinforcement learning for sharing knowledge among agents. Our ultimate goal is to drag bio-insects towards desire...Show MoreMetadata
Abstract:
In this paper, we propose fuzzy logic-based cooperative reinforcement learning for sharing knowledge among agents. Our ultimate goal is to drag bio-insects towards desired goal areas using artificial robots without any aid from human. For achieving the goal, we found an interaction mechanism using specific odor source and performed simulations and experiments in [1], [2], [3]. In the experiments, we got several problems. Due to complex and unpredictable behavior of a bio-insect it was tough to drag the bio-insect towards desired goal point. Also a huge amount of time is required to learn knowledge, and accurate knowledge to drag the bio-insect is needed. To solve the issues we designed fuzzy logic-based expertise measurement system for cooperative reinforcement learning. The structure makes artificial robots share knowledge under measuring performance evaluation of each agent. To examine the performance of the structure, we conducted experiments a number of times.
Published in: 2012 IEEE International Symposium on Intelligent Control
Date of Conference: 03-05 October 2012
Date Added to IEEE Xplore: 31 December 2012
ISBN Information: