Binary Tree and Energy Weighted Application to Three-Channel SEMG Artificial Hand

In order to solve the problem of real-time and proportional control for three-freedom artificial hand, a new method of Surface Electromyography (SEMG) disposal is proposed. This method combines binary tree with energy weighted method. By using time-domain analysis of energy, motion patterns of hand are identified by using the pattern classification of binary tree. And using multi-channel signals energy weighted to obtain coefficient which could control artificial hand to realize proportional control for multi-degree of freedom EMG artificial hand, the proportional control of electric artificial hand have been achieved according to the strength of SEMG. The experimental results indicate that the three-freedom electric artificial hand which based on electromyography has fast response and motions´ accuracy rate reaches 96%; this method is an effective way to achieve the proportional control of artificial hand.