Electroencephalogram-based control of a mobile robot

Abstract

This paper describes electroencephalogram-based control of a mobile robot. The control purpose is to achieve direction control of a mobile robot only by electroencephalogram. We develop an algorithm for detecting direction thinking ('going left' or 'going right') and apply it to direction control of a mobile robot. The detecting algorithm is based on timefrequency domain analysis using continuous wavelet transformation. Our experimental results demonstrate the possibility of achieving direction control of a mobile robot only by electroencephalogram.

Electroencephalogram‐based control of a mobile robot

“…In Section 4, the authors propose an analytical method that is robust in the face of such noise. Note that in the authors' past research [12][13][14], because the individual differences in brain waves were so significant, the authors recommended creating recognition patterns for each individual. In this research as well, the authors create recognition patterns for each individual.…”

“…The authors proposed a direction recognition system [12][13][14] using brain waves for pattern matching emphasizing the correlation coefficient between electrodes, and achieved roughly 70% accuracy rate for discriminating left and right when the eyes were closed. However, in the system the accuracy rate dropped when the eyes were opened due to the noise of eyeball movement and blinking, and so the system was considered insufficient for practical applications.…”

A Brain Computer Interface based Humanoid Robot Control System

“…The Cerebus TM Data Acquisition System acquires neural signals from 32-channels through an electrode cap on scalp. The neural signal processing section filters out high frequency noise and decomposes the filtered signals into delta, theta, alpha, and beta bands with a group of band-pass filters or wavelet filters [7]. The behavior recognition and mapping section recognizes the mental activities based on neural signal patterns to control the robot walking behavior.…”

“…First, we use a low-pass filter to filter out high-frequency noise from neural signals recorded by the EEG system and then decompose the brainwaves into delta-band [0-4 Hz], theta-band [4][5][6][7][8], alpha-band [8][9][10][11][12][13], and beta-band [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]. The decomposition can be done with a group of filters or through a Discrete Wavelet Transform (DWT) [19].…”