为了改进脑-计算机接口通信的速率,根据模拟人类自然阅读诱发模式实验数据中,噪声和信息的统计特征随时间分段变化的特点,对RLS算法进行改进,提出分段遗忘因子RLS算法.一系列的仿真和对实际实验数据的处理表明,运用我们的方法使得单次实验提取事件相关脑电位成为可能,使脑-计算机接口的通信速率得到提高.%The statistics of experimental data of imitating-natural-reading evoking mode is variable with time. For this reason, a subsection change forgetting factor RLS algorithm is proposed to extract the carrier signals from a record with the background EEG noises. Results from a series of experiments show that, single trial extraction of ERP could he realized by these processing methods, and the transfer rates of BCI would be improved.
展开▼