Mobile robot control via brain computer interface and fatigue detection based on EEG signals

Abstract

In the last decade, the rapid development of complex methods for recording brain signals and the exponential rise of available computing power as well as the increased awareness of brain dysfunctions and mental disorders, have led researchers to use large-scale neurophysiological recordings for abnormal behaviours detection, diseases diagnosis, and motor control. Electroencephalograms (EEG) are a very popular measurement for brain activities because of their non-invasive nature and their wide spectrum of possible applications. In this context, two applications have been developed in this project, the first aims to design a novel Brain Computer Interface (BCI) architecture based on Motor Imagery (MI) for real time control of a mobile robot. Spectral power computing, multi-class Common Spatial Pattern (CSP), and Machine Learning (ML) have been used to reach this aim. The second involves the proposal of an approach for fatigue detection using machine Learning (ML), Deep Learning (DL), and Genetic Algorithms (GA).