Fuzzy PD Controller in NAO System's Platform

Abstract

Humanoid robotic platforms rarely achieve the desire trajectory because of the deviation generated during the robot walking. This problem is due to different circumstances such as robot manufacturing, wear and tear of mechanic parts, or variations of floor flatness. Currently, one of the humanoid robots on the market is the robotic platform developed by Aldebaran Robotics called NAO robot, and it is used for different purposes where the robot needs to navigate into controlled spaces. NAO presents the issue of deviation during walking; therefore, a Fuzzy PD Controller is developed and implemented for this platform to reduce the orientation error and to ensure reliability during navigation. Inertial sensors are used to get the orientation reference and for feedback of the closed-loop control. Consequently, a robust control was implemented and tested in different conditions of floor and velocity during the robot’s navigation such as robot races and maze resolution. Experimental results show that fuzzy controller achieves significant improvements in the trajectories of NAO.