【摘要】 本研究提出了一种基于独立分量分析和遗传算法的诱发电位提取方法。以EP信号的峭度作为目标函数基准点,以其它信号的峭度与基准点的偏离程度的倒数作为目标函数,通过遗传算法在全局搜索适应度函数的最大值,从而首先将EP信号首先分离出来,实现了诱发电位的快速提取,而且适用于超高斯和亚高斯信号同时存在的情况。与其他提取EP信号特征的方法(如rICA)相比,本研究的方法不需要引入参考信号,只需要知道所要提取信号的峭度的估值即可,计算简单。更多还原

【Abstract】 This paper proposed a new EP extraction method based on independent component analysis(ICA) and genetic algorithm.It took kurtosis of EP as reference mark,the inverse of kurtosis differences between the other mixed signals and EP signal as object function.The overall optimal value of the fitness function was obtained with genetic algorithm.This algorithm extracted the EP signal firstly from the mixed signal.It can be used in both super-Gaussian and sub-Gaussian conditions.Compared with other ICA methods,the proposed one has the features of simple calculation and no reference signals required.更多还原