目的　以门诊量资料为例 ,计算功率谱密度函数 ,研究时间序列信号的频率域特征。方法　计算A.Schuster倡导的周期图 ,确定各周期成分的功率值 ,绘制功率谱密度函数图。结果　不同科别门诊量时序对应的频率特征得以直观表达 ,特征峰位置各有特点。经检验 ,时序为白噪声的零假设被拒绝。结论　在频率域中对时间序列进行谱分析 ,是对时间序列数据进行统计特征描述与推断的重要工具。 Objective To use outpatient volume data as an example to calculate the power spectral density function and study the frequency domain characteristics of time series signals. Methods Calculate the cycle diagram advocated by A.Schuster, determine the power value of each cycle component, and plot the power spectral density function graph. Results The frequency characteristics corresponding to the timing of outpatient visits in different departments were visually expressed, and the characteristic peak positions had their own characteristics. After testing, the null hypothesis of timing noise is rejected. Conclusion Spectrum analysis of time series in the frequency domain is an important tool for statistical description and inference of time series data.