초록 close

This paper aims to develop a wheelchair steering device for quadriplegics and arm amputees through recognizing driver’s head movement and surface electromyography (sEMG) signal. Go-and-stop motion of the wheelchair is controlled by the EMG signal produced around driver’s temple during teeth clenching, whereas left-and-right cornering motion is controlled based on the yaw angle of the head measured by a position sensor. Signal processing is designed to be robust against disturbances from road condition and to make the head movement motion distinguishable from reference coordinate of the wheelchair. Experiments were conducted 10 times and successfully discriminated clenching signals from disturbance signals on concrete pavers and speed bumps with the accuracy of 100% and 90% on low stairs. Proposed wheelchair steering interface is found to be robust against disturbances from road conditions and able to detect user’s control intention. Even though the device took twice as more time to complete a track compare to joystick, the most widely used steering device, the proposed device was found to be safe and easy to learn. The fact that the wheelchair changes the direction according to user’s head position makes this device intuitive. For these reasons, proposed device is expected to increase usability.