中文摘要：

通过对实测病人流进行非线性判别,可以为实际病人流的预测和控制提供理论指导。替代数据检验和功率谱分析是判定时间序列中是否具有非线性特性的有效方法。以时间可逆性作为非线性判据,结合功率谱分析对实测病人流的时间序列进行了非线性判别。实证结果表明,病人流在不同的时间尺度下均具有非线性特征,可以用粉红噪声刻画,并具有反持续性,及均值回复特征。

英文摘要：

Determine the characteristics of patient flows can provide a theoretical foundation for their prediction and control.Surrogate data testing and power spectral analysis are effective methods to discriminant the non-linear characteristics of time series.In this paper,the author uses the surrogate data,with the testing criterion-time reversibility,combined with the power spectral analysis to discriminant the non-linear characteristics patient flow time series.The empirical results show that the patient flows have nonlinear characteristics with different time scales,characterized by the pink noises.This indicates that the time series of patient flow is anti-persistence and a mean reversion process.