• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.65-70
• /
• 2001

We consider a hyper-redundant system that consists of many uniform units. The hyper-redundant system has many degrees of freedom and it can accomplish various tasks. Applysing the reinforcement learning to the hyper-redundant system is very attractive because it is possible to acquire various behaviors for various tasks automatically. In this paper we present a new reinforcement learning algorithm "Q-learning with propagation of motion". The algorithm is designed for the multi-agent systems that have strong connections. The proposed algorithm needs only one small Q-table even for a large scale system. So using the proposed algorithm, it is possible for the hyper-redundant system to learn the effective behavior. In this algorithm, only one leader agent learns the own behavior using its local information and the motion of the leader is propagated to another agents with time delay. The reward of the leader agent is given by using the whole system information. And the effective behavior of the leader is learned and the effective behavior of the system is acquired. We apply the proposed algorithm to a snake-like hyper-redundant robot. The necessary condition of the system to be Markov decision process is discussed. And the computer simulation of learning the locomotion is demonstrated. From the simulation results we find that the task of the locomotion of the robot to the desired point is learned and the winding motion is acquired. We can conclude that our proposed system and our analysis of the condition, that the system is Markov decision process, is valid.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2009.10a
• /
• pp.143-144
• /
• 2009

• The Journal of Korea Robotics Society
• /
• v.14 no.3
• /
• pp.196-202
• /
• 2019

Snake robots are slower than wheeled robots or legged robots, while they have an excellent terrainability in a disastrous area. Considering their advantages and disadvantages, a legged robot whose legs are snake robots, 'Quadnake' was proposed in this research. Five motions of the snake were analyzed. Applying these motions, Quadnake could implement eight kinds of motions which snake robots and quadruped walking robots can implement. As a result of it, Quadnake can have the advantages of both a snake robot and a walking robot. It is expected to move stably in a harsh terrain with snake's motion and move fast with walking.

• Proceedings of the KIEE Conference
• /
• 2003.11b
• /
• pp.7-10
• /
• 2003

뱀형 로봇은 자유도보다 액추에이터의 수가 적은 논홀로믹 구속조건(nonholonomic constraint)을 가지며, 단순한 신체구조 이지만 초-여유자유도 구속조건(hyper-redundant constraint)을 이용해서 기밀한 운동과 다양한 기능을 만들어내는 특징을 가지고 있다. 본 논문에서는 6개의 관절로 각 링크가 2차원 상에서 직렬로 연결된 뱀형 로봇의 기구설계 및 기구학과 동력학을 바탕으로 설계된 기구에 대해 해석하여 운동방정식을 유도하여 추진원리와 운동원리에 관하여 알아본다. 기본적인 운동 메커니즘을 해석하여 구현한 알고리즘을 제작한 로봇에 적용하여 추진 원리와 운동원리를 검증한다. 실험용 로봇은 링크 중앙에 법선 방향으로 마찰력이 발생할 수 있도록 수동바퀴를 가지고 있으며, PC와 RF(Radio Frequency)로 직렬통신을 하며 PC에서의 운동 명령의 조작에 의해 전진, 후진, 좌/우 방향으로 회전을 할 수 있도록 운동 알고리즘을 적용할 수 있도록 제작되었다. 특징으로는 일반적으로 토크를 입력으로 하지 않고 각도를 입력으로 하여 관절을 제어하고 있다는 점이 있으며, 운동방정식 또한 이에 대한 관계를 바탕으로 유도한 것이다.