摘要
提出了一种结合AR模型和贝叶斯分类的肌电信号动作模式识别方法。首先将采集到的肌电信号进行预处理,提取AR系数作为其特征值。其次设计了一个贝叶斯分类器,利用AR系数对手臂的各种肢体动作进行动作模式分类。实验表明这种方法不仅降低了误识别率,而且取得了比较理想的识别效果。同时,采用虚拟仪器技术提高仪器的测量精度,降低成本,降低计算工作量。
In this paper, a method to analyse surface myoelectrical signal （SMES） is provided, which based on AR model and Bayes taxonomy. First, we use AR model to predispose original signal, distill the calculated AR coefficients as its eigenvector. Second, a Bayes statistics algorithm is designed to recognize the pattern movement of forearm muscle with AR coefficients. The experiment indicates this measure can reduce the rate of error identification and the results of pattern recognition are relative well. It also proves that there are some relations between motion patterns and AR coefficients. At the same time, adopt virtual instrument technology to raise accuracy of measurement, reduce the cost and workload.
出处
《杭州电子科技大学学报（自然科学版）》
2005年第5期14-17,共4页
Journal of Hangzhou Dianzi University：Natural Sciences
基金
2005年度浙江省高校青年教师资助计划项目
浙江省自然科学基金资助项目(RC02070)
关键词
自回归模型
表面肌电信号
贝叶斯分类
虚拟仪器
AR model
surface myoelectrical signal （SMES）
Bayes taxonomy
virtual instrument
