Modifiable walking pattern generation for humanoid robot using allowable ZMP variation

Abstract

In order to handle complex navigational commands, this paper proposes a novel walking pattern generation algorithm for a humanoid robot, which can modify a walking period, a step length and a walking direction in both sagittal and lateral planes. By allowing a variation of ZMP over the convex hull of foot polygon, it is possible to change the center of mass (CM) position and velocity independently throughout the single support phase without any extra step for adjusting the CM motion. This permits a range of dynamic walking motion, which is not achievable using the 3D Linear Inverted Pendulum Mode (3D-LIPM). In addition, the proposed algorithm enables to determine the dynamic feasibility of desired motion via the construction of feasible region, which is explicitly computed from the current CM state with simple ZMP functions. Moreover, adopting the closed form functions makes it possible to reduce the computational cost, which ensures a real-time calculation. Also, to stabilize the robot in the presence of external disturbance, stabilization strategy is proposed, which is composed of two stages: Disturbance absorbing stage and state recovery stage. In the disturbance absorbing stage, foot impedance was relaxed to keep the contact condition. Also, foot position was modified to avoid the early-contact. In the state recovery stage, foot impedance was returned to the normal state and the CM motion was modified to arrive at an appropriate state during double support phase. The effectiveness of the proposed algorithms are demonstrated through both computer simulation and experiment on the humanoid robot, HanSaRam-VII and VIII, developed at RIT laboratory, KAIST