GAを用いた多足歩行ロボット制御

-ROSによるシミュレーション-

抄録

In this paper, it is shown about the walking pattern making method of tideland research robot. This method used GA (Genetic Algorithm). The tidelands are special environments. It has water areas and land areas. So, the research robot must be able to move in those two environments. We propose the multi legged walking robot as a model. This can freely move in tidelands. This robot has 6 legs and 6 joints. So that, it does the tripod gait like ants on the land. And, this robot swims to use the fin motion in the water. The multi legged walking robot can move according to the surrounding environment. We considered about the control method using GA. In this method, the robot is walked using angular velocity control. GA is used to create parameters necessary for angular velocity control. We confirmed that it can make the stable walking pattern by this method.