Standing and Stepping Control with Switching Rules for Bipedal Robots Based on Angular Momentum Around Ankle

Abstract

In this paper, we propose standing and stepping control with switching rules based on angular momentum around the ankle for planar bipedal robots. A theoretical analysis under some approximation and mass distribution conditions shows that the proposed standing control maximizes stable regions. We can then classify the state of robots into the following three categories: (1) stabilizable via ankle torque; (2) unstabilizable only via ankle torque and stabilizable via ankle torque and trunk posture control; and (3) unstabilizable via ankle torque and trunk posture control. This criterion enables switching rules to appropriately switch robot control to balance control via ankle torque, balance control via ankle torque and trunk posture control, or stepping control. The proposed method is applicable to robots without feet. Simulation results demonstrate that the proposed method appropriately switches control according to the amplitudes of disturbances and maintains the balance of robots with and without feet.