混沌时间序列法在地下水位预测中的应用

Application of Chaos Time Series Method in Groundwater Table Prediction

在相空间重构的基础上,并借助G-P(Grassberger-Procaccia)算法、C-C方法和Wolf方法分别从濮阳市范县、华龙区和南乐县中的某观测孔的地下水位一维时间序列中提取lyapunov指数,通过对地下水时间序列的混沌特性识别,表明其时间序列具有混沌特征。然后运用加权一阶局域法对地下水位时间序列进行预测。结果表明,该模型用于地下水位时间序列的短期预测是可行和有效的。

Applying G-P （Grassberger-Procaccia）arithmetic, C-C arithmetic （the improved method） and Wolf method based on phase space reconstruction method, Lyapunov exponents are distilled from one-dimension time series of underground water table in Fanxian County, Hualong District and NaMe County of Puyang City. The results indicate that this time series possesses the character of chaos. Then, an adding-weight one-rank local-region forecasting model is developed for the prediction of underground water table. The results show that the model is feasible and valid for the short term prediction of the time series of underground water table.