Research of BCI system based on neural networks with high generalization and multi-channel orthogonal components

开发了脑机接口（BCI）系统。提出了一种强调脑电波特征的方法。采用正交分量和并行多层神经网络对心理任务进行分类。将这些结果求平均值以得到最终结果。该方法在心理任务分类方面取得了较好的效果。另一种将特殊滤波器、二值分类器和纠错码相结合的方法获得了较高的性能。

Brain Computer Interface(BCI) system has been developed. A method to emphasize features of the brain waves has been proposed. The orthogonalized components and parallel multi-layer neural networks are used to classify themental tasks. These results are averaged to obtain the final result. The high performance for mental task classification has been obtained. Another method, combining the special filter, binary classifiers and error correcting code, has beenproposed, resulting in high performance.