Biomimetic Locomotion Control of a Quadruped Walking Robot ⋆ Ig Mo Koo ∗ Tran Duc Trong ∗ GiaLoc Vo ∗ Young Kuk Song ∗ and Hyouk Ryeol Choi ∗ Tae Hun Kang ∗∗ ∗ Sungkyunkwan university, Chunchun-dong 300, Kyunggi-do, Korea(Tel: +82-31-290-7449; e-mail: hrchoi@me.skku.ac.kr). ∗∗ Phohang Institute of Intelligent Robotics, Korea(e-mail: bigxihn@postech.ac.kr) Abstract: In this paper, a new biomimetic control method for a quadruped walking robot is proposed. The method is derived by the observation of the gravity load receptor and stimulus- reaction mechanism of quadrupeds’ locomotion, and the study of the stances on walking and energy efficiency. Though the controller is simple, it provides a useful framework for controlling a quadruped walking robot. In particular, by introducing a new rhythmic pattern generator the heavy computational burden to be paid on solving kinematics is relieved. The effectiveness of the proposed method is validated via a dynamic simulation and experimental works in a quadruped walking robot, called AiDIN(Artificial Digitigrade for Natural Environment). 1. INTRODUCTION In spite of rapid development in robotic technologies, liv- ing creatures are still superior to robots existing currently in various aspects. Thus, it is necessary to understand the principles underlying the motions and behaviors of biological subjects for the control of robots. Mimicking living creatures currently becomes one of the worldwide trends for robotic innovations. It is considered as one of the most adequate way of developing a robot since biological systems provide a number of useful ideas concerning the control of robots. Recently, robotic researchers as well as biologists propose innovative ideas for the control of the walking robot system. Among several ideas, mimicking the rhythmic motion of animals is one of the most promising ways to control the walking robot system. By studying on this, the locomotion of the walking robot can be close to that of the real animal. According to recent studies of neurobiologists, it is noted that a part of the neural system in the brain of animals, called Central Pattern Generator(abbreviated as CPG), produces rhythmic movements in the locomotion. Since CPG can generate rhythmic outputs even without the sensory feedback, it has been one of the most attractive approaches in the biomimetic control. On the way to study the CPG, many neural models such as McCulloch and Pitts neuron, Leaky integrator neuron, and Matsuoka neuron have been proposed by investigating neurons in the real animals’ brains or bodies[1, 2]. As efforts of developing relevant methods to control robots, algorithms inspired from rhythmic patterns generators have been studied[3]. Though these neural models can illustrate the control method of animal movements, the results of applications in robotic systems are far different from real biological ⋆ This research was supported in part by the project of the dual- use technology for military and civilian missions(”Development of Quadruped Robots”) of the Ministry of Commerce, Industry and Energy (MOCIE), KOREA. Motor driver DC Motor Clutch Spring SBC(Windows-XP) SBC(RTLinux) CCD Speaker Force sensor (hidden) Gyroscope (hidden) Fig. 1. AiDIN(Artificial Digitigrade for Natural Environ- ment) systems because animal’s movement can not be simply mimicked with several neurons that robotic systems have introduced until now[3]. On controlling a quadruped walking robot, various aspects should be taken into account. In this research, several crit- ical considerations are given for the design of a quadruped walking robot controller. Mainly, the basic principles of the quadrupedal locomotion controller are analyzed in terms of a biomimetic point of view. By the observation of the gravity load receptor, stimulus-reaction mechanism of animals are investigated using the stances on walking and posture control. The controller design factors are feedback information from several receptors and stimulus-reaction based on biological sensory system analysis. The proposed method is validated via dynamic simulation using Open Dynamic Engine(abbreviated as ODE) and then, it is Proceedings of the 17th World Congress The International Federation of Automatic Control Seoul, Korea, July 6-11, 2008 978-1-1234-7890-2/08/$20.00 © 2008 IFAC 3011 10.3182/20080706-5-KR-1001.4100