基于肌电信号产生机理 ,对双通道前臂肌电信号建立单输入多输出 IIR系统模型 ,由于模型输入未知且不可测 ,采用了盲信道辨识方法对模型传递函数进行辨识 .通过提取模型参数作为信号特征 ,能够对握拳、展拳、前臂内旋和前臂外旋四类前臂动作进行识别 .实验表明 ,该方法运算量小 ,适合在线实现 ,性能要优于传统的 AR模型方法 Based on the generation mechanism of EMG signals, a single-input multiple-output IIR system model is established for the two-channel forearm EMG signals. Blind channel identification is used to identify the model transfer function because the model input is unknown and unmeasurable. By extracting the model parameters as signals Which can identify four kinds of forearm movements such as fist, flexing, forearm pronation and forearm pronation.Experiments show that this method is less computationally intensive and suitable for on-line implementation, and its performance is better than the traditional AR model