运用不稳定周期轨道方法识别神经元阵发放电序列中的确定性动力学

Identification of determinant dynamics in neurons spiking series by using unstable periodic orbits

目的 研究如何识别神经元阵发放电序列中的确定性动力学 .方法 运用不稳定周期轨道及关联分维、非线性预报对神经元阵发放电序列中的确定性动力学进行了研究 .结果 在确定性统计方法失效的情况下 ,不稳定周期轨道方法依然能可靠的辨别出放电序列的周期 ,周期 ,周期 轨道结构 .结论 与确定性统计方法比较 ,不稳定周期轨道更适合于对高维动力系统的混沌时间序列进行分析 .

AIM To study how to identify determinant dynamics in neurons spiking series. METHODS Determinant dynamics in neurons spiking series was researched by using unstable periodic orbits, correlation dimension and nonlinear prediction. RESULTS When the determinant statistics is ineffetive,the orbital structures of Period Ⅰ, Period Ⅱ and Period Ⅲ could be recognized by using unstable periodic orbits. CONCLUSION Compared with determinant statistics, unstable periodic orbits is more suitable for the analysis of chaotic time series with high dimension.