Resistant Compliance Control for Biped Robot Inspired by Humanlike Behavior
Qiang Huang,Chencheng Dong,Zhangguo Yu,Xuechao Chen,Qingqing Li,Huanzhong Chen,Huaxin Liu
DOI: https://doi.org/10.1109/TMECH.2021.3139332
2022-01-01
IEEE/ASME Transactions on Mechatronics
Abstract:Compliance control is important for the realization of disturbance absorption in biped robots. However, under a sustained disturbance, compliance control causes the robot's balance to deteriorate because of its floating base nature. Humans address this problem by resisting external disturbance. When pushed, a human will reconcile their posture with the applied external force and then push back to maintain balance. Inspired by this behavior, we propose a compliance control strategy for biped robots called resistant compliance, which allows a robot to comply with the external disturbance initially and then repel the disturbance to reduce the imbalance caused by the reconciliatory motion. As a result, the robot can obtain improved environment-interaction stability and react more like a typical human, thus making both its locomotion and its interactions more stable and safer. To realize this control strategy, the virtual-mass-model (VMM) control is redesigned to unify the disturbances from unexpected external forces and an inclined floor. Then, the VMM control is combined with the linear-inverted-pendulum model to realize resistant compliant motion. Model predictive control is used to track the reference zero-moment-point trajectory, which is essential for locomotion. To validate the proposed control strategy, the method is implemented on the human-sized humanoid robot BHR-T.