EMG sliding mode finger joint synergy control of a dexterous artificial hand

Abstract

From observation of human data, a set of sinusoidal trajectories were developed to mimic the human motion of unscrewing a bottle cap with the thumb and index finger. These trajectories were then implemented on a dexterous robotic hand in the form of a robustly stable sliding mode control algorithm. With the developed synergies, a single myoelectric input was used to control multiple finger joints simultaneously. This controller was then compared to a one degree of freedom prosthetic hand also under sliding mode control. The two hands were used to complete an unscrewing task as quickly as possible. The synergy controller produced a statistically significant reduction in task completion time and also reduced the required workspace.