中文摘要：

根据癫痫发作过程中，EEG信号表现出来的发作间期和发作期2种不同的状态，通过分析发现在该过程中大脑动力系统存在不同的动力学嵌入空间，存在不同的吸引子。还应用伪邻点法、互信息法和C—C方法进行了推导和仿真，对2种不同状态进行相空间重构，确定了癫痫病人不同状态EEG不同吸引子的参数，并在此基础上提出了若干新的见解。

英文摘要：

During the process of epileptic seizure, there are two states of the epileptic brain： the interictal state, steadystate electroencephalography （EEG） ongoing activity, and the ictal state, which is characterized by paroxysmal occurrence of synchronous oscillations. Different dynamic phase spaces and chaotic attractors are found during the process. By using methods of False Nearest Neighbors, mutual information and C -C method, phase space is reconstructed and parameters of chaotic attractor are computed and some new viewpoints on this foundation are put forward.