In the second case, we assume that the elbow and forearm
torques cannot be estimated based on EMG signals. Therefore,
the robot estimates the torques of elbow and forearm based on
EEG signals. The estimation result is shown in Fig. 9. In Fig. 9,
the black solid line is the calculated result based on measured
joint angles, and the red broken line is the result which is
estimated based on EMG and EEG signals (EMG and EMG
case). Each axis is shown in Fig. 6. In addition, Table III shows
the error values. One is the error value of each axis between the
measured result and estimated result based on 16 EMG signals.
Another is the error value of each axis between the measured
result and estimated result based on EMG and EEG signals.
Although the accuracy of the estimation results with EEG and
EMG signals is lower than that with only EMG signals, we
could estimate the hand velocity based on EMG and EEG
signals as shown in Fig. 9 and Table III.