TY - JOUR
T1 - Robot Simulations Based on Bipedal Spring-Mass Model With Variable Slack Length and Stiffness
AU - Xu, Zhe
AU - Gao, Junyao
N1 - Publisher Copyright:
© 2013 IEEE.
PY - 2017
Y1 - 2017
N2 - This paper proposed an extension of bipedal spring-mass model with variable slack length and stiffness. Since conventional bipedal spring-mass model has problems to accomplish stepping and low speed walking, we solve it by adjusting slack length of spring leg. Moreover, to produce heel-strike and toe-strike running, the stiffness of spring is changed to get running gaits with different frequencies. In order to imitate the human foot rolling behavior, we synchronize motion of center of pressure with that of center of mass. Based on the proposed methods, simulations of a human size robot achieve walking, heel-strike, and toe-strike running. Walking at a speed of 1 m/s can generate the double-humped ground reaction force. Heel-strike and toe-strike running gaits at a speed of 4.5 m/s are accomplished. The frequencies of two running gaits are consistent with those in human motion capture data. Joints angular positions of walking and heel-strike running are similar to those of human data. Therefore, the proposed model turns out to be effective in generating human-like walking and running behaviors.
AB - This paper proposed an extension of bipedal spring-mass model with variable slack length and stiffness. Since conventional bipedal spring-mass model has problems to accomplish stepping and low speed walking, we solve it by adjusting slack length of spring leg. Moreover, to produce heel-strike and toe-strike running, the stiffness of spring is changed to get running gaits with different frequencies. In order to imitate the human foot rolling behavior, we synchronize motion of center of pressure with that of center of mass. Based on the proposed methods, simulations of a human size robot achieve walking, heel-strike, and toe-strike running. Walking at a speed of 1 m/s can generate the double-humped ground reaction force. Heel-strike and toe-strike running gaits at a speed of 4.5 m/s are accomplished. The frequencies of two running gaits are consistent with those in human motion capture data. Joints angular positions of walking and heel-strike running are similar to those of human data. Therefore, the proposed model turns out to be effective in generating human-like walking and running behaviors.
KW - Center of pressure
KW - bipedal spring-mass model
KW - variable leg length
KW - variable leg stiffness
UR - http://www.scopus.com/inward/record.url?scp=85018528387&partnerID=8YFLogxK
U2 - 10.1109/ACCESS.2017.2652128
DO - 10.1109/ACCESS.2017.2652128
M3 - Article
AN - SCOPUS:85018528387
SN - 2169-3536
VL - 5
SP - 1156
EP - 1163
JO - IEEE Access
JF - IEEE Access
M1 - 7814262
ER -