Energy-Efficient Walking for Biped Robot Using Self-Excited Mechanism and Optimal Trajectory Planning

Recently, a lot of researches which aim at realization of dynamic biped walking are being performed. There is Honda's ASIMO as a representative of researches of humanoid robots. ASIMO has joints of many degrees of freedom that are near to a human being, high environment adaptability and robustness, and can do various performances. However, it needs many sensors, complicated control, and walks with bending a knee joint to keep the position of a centre of gravity constant. Therefore, it walks unnaturally and consumes much energy. On the other hand, McGeer performed the research of passive dynamic walking from the aspect of that it is natural motion in a gravitational field (McGeer, T., 1990). This robot which could go down incline only using potential energy was developed and realized the energy-efficient walking. However, it needs incline, and its applied range is small because it has no actuator. Therefore, the researches that aimed at energy-efficient biped walking on level ground have been performed. S.H.Collins exploited the robot which had actuators at only ankles (Collins, S. H. & Ruina A., 2005). M.Wisse exploited the robot which used pneumatic actuators (Wisse, M. & Frankenhuyzen, J. van, 2003). Ono exploited the self-excitation drive type robot which had an actuator only at hip joint (Ono, K. et al, 2001); (Ono, K. et al, 2004); (Kaneko, Y. & Ono K., 2006). And then, Osuka and Asano performed the level ground walking from a point of view to mechanical energy for joints which is the same with the energy consumed of passiveness walk (Osuka, K. et al, 2004); (Asano, F. et al, 2004). These biped robot's studies used the technique of the passive dynamic walking which used inertia and gravity positively by decreasing the number of actuators as much as possible. However, in order to adapt the unknown ground, the biped robot needs actuators to improve the environment adaptability and robustness. Here, Ono proposed the optimal trajectory planning method based on a function approximation method to realize an energy-efficient walking of the biped robot with actuators similar to a passive dynamic walking (Imadu, A. & Ono, K. 1998); (Ono, K. & Liu, R., 2001); (Ono, K. & Liu, R., 2002); (Peng, C. & Ono K., 2003). Furthermore, Huang and Hase verified the optimal trajectory planning method for energy-efficient biped walking by experiment, and proposed the inequality state constraint to obtain better solution which is desirable posture in the intermediate time of the walking period (Hase, T. & Huang, Q., 2005); (Huang, Q. & Hase, T., 2006); (Hase, T., et al., 2006).