中文摘要：

为了提高人体手部运动模式识别的准确性,提出了一种基于人工鱼群算法优化支持向量机（SVM）的模式识别方法。该方法对采集的表面肌电信号（s EMG）去噪后提取小波系数最大值作为特征样本,将提取后的特征输入到SVM进行动作模式识别,同时采用人工鱼群算法优化SVM（AFSVM）的惩罚参数和核函数参数,避免参数选择的盲目性,提高模型的识别精度。通过对内翻、外翻、握拳、展拳四种动作仿真结果表明：该方法与传统的SVM方法相比具有更高的识别率。

英文摘要：

To improve accuracy of human hand motion pattern recognition,a pattern recognition method for optimizing support vector machine（ SVM） by using artificial fish swarm algorithm（ AFSA） is proposed. The maximum value of wavelet coefficients is extracted as feature samples from the de-noised surface electromyography（ s EMG） signals,then the extracted feature is inputted into a SVM to classify actions recognition,and at the same time,AFSA is used to optimize the penalty parameters and the kernel parameters of the SVM,which avoids the blindness of parameters selection and improves recognition precision of the model. Simulation results show that four movements（ wrist down,wrist up,hand grasps,hand extension） are successfully identified by the method of SVM combined with AFSA. Compared with the traditional SVM method,the method has higher recognition accuracy.