移動目標物の捕獲問題におけるいくつかの協調行動の学習法

概要

論文の詳細を見る
In this paper, some learning methods of cooperative behaviors for a group of mobile agents(robots) that capture fleeing targets are presented.Specifically, the problem that two agents capture a fleeing target, is called the pursuit problem.It is assumed that the system is composed of autonomous mobile agents of the same type and that each agent behaves according to its own transition rules.Transition rules to capture targets are encoded as the genes of each agent and GA applies genetic cooperation to those genes.Transition rules are determined as a function of (1) assignment of agents and targets, (2) sums of product of weights and unit vectors for coordinates of agents and targets.And (3) Transition rules using only weights depending on distance as the special case of (2), are used.Simulation results show that the number of the targets captured by the robots increases as the time passes and, therefore, a cooperative behavior is learned by all the proposed systems.Specifically, it is shown that the model with the condition (1) is the best one in the presented problems.
社団法人電気学会の論文
2003-03-01