Self-Stabilizing Dynamics for a Quadruped Robot and Extension Toward Running on Rough Terrain

Abstract

We designed and analyzed a control strategy that achieves autonomous adaptation and good energy efficiency in running by a quadruped robot. Our control strategy, inspired by previous studies on self-stabilizing dynamics, combines rhythm and torque generators with delayed feedback control (DFC) to achieve stable running and essential energy input. We developed an adaptation strategy to extend this control strategy that adjusts the robot’s leg touchdown angle based on the body’s pitch angle. Used together with our proposed control, it enables robust bounding over a shallow slope. Simulation results confirmed the feasibility of our proposal and its performance.