中文摘要：

仿生肌电假肢的控制依赖于表面肌电信号,其中基于表面肌电信号的肢体动作识别是关键。该文提出一种时域波幅直方图和频域功率谱比值相结合的特征提取方法,通过对两路表面肌电信号的波幅直方图分析和频谱分析,以波幅比值作为时域特征,以功率谱比值作为频域特征,通过证据理论对它们各自利用神经网络得到的分类结果进行证据累积,最终得到分类结果。实验证实了该算法的识别率高于利用单一特征参数的分类方法。

英文摘要：

To improve the identify ratio of the surface electromyography(SEMG),this paper extracted the SEMG parameters on the time domain and the frequency domain respectively,and completed the hand movement recognition by using the support vector machine.The amplitude ratio is the parameter on the time domain through the analysis of the SEMG amplitude histogram,and the power spectrum ratio is the other one on the frequency domain.Then the outputs of the BP networks whose inputs are the amplitude ratio and the power ...