【摘要】 肌电信号是与神经肌肉活动有关的生物电的体现 ,其应用已逐步深入到生物医学的各个领域。肌电信号模式识别是肌电应用的基础。本文在表面肌电模式识别中引入了特征优化环节 ,并讨论了其作用。通过对 7个志愿者的实验研究表明 ,采用特征优化情况下 ,动作平均辨识率可提高近 12 %。结论表明 ,在没有增加原始信息的情况下 ,通过合适的特征优化处理可提高特征的可分性 ,从而提高动作模式的辨识率。这说明了特征优化在肌电信号模式分类中有重要作用 ,对其在康复控制等实际应用场合也会产生积极影响。更多还原

【Abstract】 Electromyography (EMG) is a bio\|electrical manifestation related to neuromuscular activities. It has been used widely in various fields of biomedicine, where pattern recognition plays an important role. Feature optimization was introduced in the pattern recognition of surface electromyography and is presented in this paper, and its effects are discussed in detail. With experiments from seven volunteers, it showed that the average recognition rate of the motions can improve by 12%. It alss shows that, without new original information, the separability of features could be improved with appropriate optimization method, so that the probablity to recognize patterns of motions can be improved. The feature optimization does play important role in the process of pattern recognition, which would promote practical applications, such as the fielde of rehabilitation etc.更多还原