Modeling the THBIP-1 Humanoid Robot Control System

Abstract

The paper presents a gait planning method and control strategy for THBIP-1 (Tsinghua Biped Humanoid Robot) to realize stable gait. Our research was carried out in three directions: the study of human walking properties by energy analysis, the development of an experimental humanoid robot apparatus, and the study of a natural dynamic gait and balance control scheme. The main goal of creating a humanoid robot is to replace humans in jobs that are dangerous or repetitive for humans. This robot includes a head, torso, two arms, two legs, totaling 32 DOFs (degrees of freedom). This project aims to realize stable walking in different environments for an autonomous humanoid robot. First, design issues including the robot's mechanical structure and control architecture are defined. Second, a gait generation method based on the optimization of the main support leg is presented, and a control strategy consisting of a local joint controller and a sensor feedback controller is also described. As a result of the analysis of the robot's walking experiments, it was determined that the robot has the ability to walk stably on the ground, rotate in any direction, and climb up/down the stairs.