基于EEMD的公路客流量波动性与周期性研究

The volatility and periodicity of highway passenger volume based on ensemble empirical mode decomposition

为实现对公路客流量进行快速且准确预测,提出了将公路客流短期预测的总体平均经验模态加以分解的EEMD方法.通过将原始时间序列分解为多个固有模态函数与趋势项之和,对多个固有模态函数进行周期性和波动性分析,揭示出各阶模态间的周期性变化规律以及不同时间段内的波动特点,进而证明了总体平均经验模态分解方法的合理性与有效性.

In order to achieve the forecasting of the highway passenger volume quickly and accurately,the volatility and periodicity of the highway passenger volume is studied based on ensemble empirical mode decomposition（EEMD）method,which is very important for establishing and optimizing the passenger highway operation plan.The original time series is decomposed into a series of intrinsic mode functions（IMF）and a trend item.By analyzing the volatility and periodicity of each function,the different volatility features of the highway passenger volume with different scales are shown,which implies the volatility characteristics of the passenger volume in different period and the periodic property in different IMF.Consequently,it can be proved that the rationality of the EEMD method is feasible and effective.