An experimental system identification modeling and robust control for NAO humanoid robot

Using Bhuman walking engine, this paper presents an LTI model for Nao robot walking control. The robot is assumed as a point mass in a 2-D maneuver. Model calculations and validation have been done using experimental data, performing various tests on the robot, and using system identification tools. Afterwards, a robust controller is designed to control the robot position in the horizontal (x-y) plane and its performance has been compared with a well-designed PID controller. The obtained results reveal the reliability of the controller in the presence of considerable disturbances.