Departmental Bulletin Paper 生体信号収集ヘッドセットを用いた電動車いす制御システムの操作性改善

松澤, 海良

In our previous study, an electric wheelchair was controlled by the user's facial expression and head movement using the control system based on a brain-computer interface (BCI) headset having 14 measuring electrodes and a 2-axis gyroscope sensor. However, the operability issue caused by the time lag between theinput of the command and the beginning of the movement of the electric wheelchair was observed. In this paper, the control system based on the BCI headset is improved to solve this issue. In the improved control system, the user’s facial expression is distinguished via a neural network and frequency analysis withbiosignals detected by electrodes of the BCI headset. Three kinds of methods to distinguish the facial expression were newly proposed. On the other hand, the user's head movement is recognized based on threshold criteria for angular velocity and displacement detected by the gyroscope sensor. With the aim of evaluating the improved control system, an experiment to measure the distinction rate for operation commands and a video analysis to measure the time lag were conducted, and the experimental results were compared between the proposed three distinction methods and the previous method.Key Words : biosignal, Brain-Computer Interface (BCI), electric wheelchair, neural network, Fouriertransform, wavelet transform

Number of accesses :  

Other information