Motor imagery electroencephalogram signal feature extraction method

The invention discloses a motor imagery electroencephalogram signal feature extraction method, which aims to solve the problems that electroencephalogram signals have a large amount of cognitive related information but have the defects of weak information, strong background noise, difficulty in feature extraction and the like, the electroencephalogram signals are non-stationary random signals, neural activity states represented by the signals are difficult to find from the signals, and difficulty is caused for motor imagery instruction recognition. The calculation complexity and the convergence speed are balanced, implementation in a single-chip microcomputer is facilitated, and the requirement for collecting the electroencephalogram signals in real time can be met. According to the method, the convergence speed is high, the waveform shape is not easy to change, and physiological artifacts and circuit noise can be effectively removed. 本发明公开了一种运动想象脑电信号特征提取方法，脑电信号具有大量认知相关信息，但是其具有弱、背景噪声强、特征难以提取等缺点，脑电为非平稳随机信号,从信号