Abstract

The main objective of this project is to develop a brain-actuated humanoid robot navigation system for translating the various human intentions into appropriate motion commands for robotic applications. This robot navigation system uses an EEG-BCI.

Description

The experimental procedures include online feedback test sessions, real-time control sessions and the offline training sessions. During the offline training sessions, the informative feature components were selected using the linear discriminant analysis (LDA) distance metric and the Fisher ratio. By using band power analysis, the amplitude features from the EEGs are extracted in the offline training sessions.

To build an asynchronous BCI system, the Intentional Activity Classifier (IAC) and the Motor Direction Classifier (MDC) were hierarchically structured and trained. During the navigation experiments, the subject controls the humanoid robot in an indoor maze with the help of the BCI system with real-time images from the camera on the robot’s head.

Result

By using the proposed brain-actuated humanoid robot navigation system, three subjects can successfully navigate the indoor maze.