強化学習モデルを用いた掃除ロボット経路計画のためのモデルベース解析

抄録

Due to the high development of artificial intelligence technology, robotic technology and automatization of household chores or building maintenance, our living quality has been dramatically improved. It is expected that introduction of cleaning robots to indoor is more rapidly proceed than ever by this trend. An automatic cleaning robot that runs around indoors and performs cleaning, is representative as service robots. Hence we would like to introduce this paper to present the first step of intention to approach of modifcations and improvements of current cleaning robot, by applying the Reinforcement Learning algorithm to cleaning robot, making the robot to learn how to map situations to actions, so as to maximize a numerical signal (reward). Applying model-base to cleaning robot of certain area that needs to be vaccumed, the robot will achieve self-guided of planning path, to search and vacuum desired space of dirts automatically.