Abstract
This paper deals with an attempt at predicting the walking period of the next step by using myoelectrogram (EMG). In order to control an above-knee prosthesis that can change its damping coefficient of a knee-joint according to walking period on level walking, walking period must be predicted precisely during the last walking-cycle. Hence, an attempt at predicting the walking period of the next step by using EMG picked up at the timing of Heel-ON, has been made. The surface differential electrode with pre-amplifier and filtering circuit, with high-CMRR and low-power consumption was developed. The best EMG data-composing method to transmit the walking period-information was developed. The walking period prediction algorism to predict the walking period of the next step was designed, and its effectiveness was confirmed with amputee's walking experiments.
