论文部分内容阅读
三峡库区入库量作为随时间变化的序列具有非线性和非平稳性特征,所涵盖的频率信息不仅能够体现长江流域年变化尺度以上的周期特点,而且在每年汛期与枯水期的频率变化也有所不同。主要利用HHT方法对三峡库区入库量信号进行分析处理,首先利用EMD方法对信号进行分解,分析结果能够清晰分辨出由高到低不同频率范围的固有模态函数分量;对固有模态函数分量做Hilbert变换得到时频特征表明:在20d至1a的周期范围内,频率幅值变化不大,但自2003年蓄水开始,频率存在略有增大的趋势,并且洪水期与枯水期频率变化分界不明显;年周期是20a库区信号中稳定的周期性变化信息,而残差信号代表了入库量的整体趋势,反应了整个库区水量自1998年起的减小趋势。
The storage volume in the Three Gorges Reservoir area is nonlinear and non-stationary as a series of time-varying sequences. The frequency information covered can not only reflect the periodic characteristics above the annual variation scale in the Yangtze River Basin, but also change in the annual flood season and the dry season different. In this paper, the HHT method is used to analyze and process the storage volume in the Three Gorges reservoir area. Firstly, EMD method is used to decompose the signals. The analysis results can clearly distinguish the intrinsic mode function components in different frequency ranges from high to low. The Hilbert transform is used to obtain the time-frequency characteristics. The results show that the amplitude of the frequency does not change much in the period from 20d to 1a, but the frequency has a slight increase since the beginning of the impoundment in 2003, and the frequency of flood and dry season The boundary is not obvious. The annual cycle is a stable periodic change in the signal of reservoir area 20a, while the residual signal represents the overall trend of storage volume, reflecting the decreasing trend of water volume in the whole reservoir area since 1998.