Planar Biped Walking With an Equilibrium Point Controller and State Machines

Abstract

In this paper, we present both simulation analysis and experimental study of a planar dynamic biped walking robot. The proposed control structure involves an equilibrium point controller at the local joint level and state machines at the interjoint level. The robot has actuated hip joints and knee joints as well as unactuated ankle joints with curved feet. We first show in simulation analysis that stable walking by this robot is possible with various gait patterns and a wide range of walking speed. Then, we implement the controller on a real biped robot, and show in real-time experiments that, by directly changing two controller parameters, the robot can change its walking gait/speed on the fly and can achieve a very fast walking speed.