中文摘要：

针对电动假手的仿生控制问题，给出了一种三自由度实时比例控制肌电假手的设计方案。通过采集残臂上的4路表面肌电信号（SEMG），采用能量时域分析法对信号进行了特征提取，并采用二叉决策树模式分类方法识别得到了手部3个自由度7个动作模式。由单片机构成的信号处理和控制电路，实现了三自由度电动假手的实时仿生控制；并根据表面肌电信号的强弱，采用多路SEMG能量加权法来求取多自由度假手的比例控制系数，实现了对电动假手的比例控制。试验结果表明，基于表面肌电信号的三自由度肌电假手响应速度快。动作准确率达到96％以上。

英文摘要：

Aiming at the problem of electric artificial hand＇s bionic control, a design of three-freedom artificial hand with real-time and proportional control was proposed. Four-channel surface electromyography （SEMG） signals were acquired from the amputee stump, then, time-domain analysis of energy was used for the feature extraction of signals, and seven motion patterns of hand were identified by using the pattern classification of binary tree. The signal processing and control circuit which was composed of single-chip computer achieved the real-time bionic control of three-freedom electric artificial hand. And the multi-channel signals energy weighted was used to obtain coefficient, the proportional control for multi-degree of freedom EMG prosthetic hand was realized, the proportional control of electric artificial hand was achieved according to the strength of SEMG. The experimental results indicate that the three-freedom electric artificial hand based on electromyography has fast response and motions＇ accuracy rate reaches 96%.