基于混沌时间序列的地下水动态预报  

Prediction of underground water based on chaotic time series

作　　者：蔡明科[1] 胡国杰[1] 许义和[1] 

机构地区：[1]西北农林科技大学水利与建筑工程学院,陕西杨凌712100

出　　处：《水资源与水工程学报》2010年第5期106-108,112,共4页Journal of Water Resources and Water Engineering

基　　金：国家自然科学基金项目(50879071);水利部公益性行业科研基金项目(200801104)

摘　　要：以重构相空间理论为基础,介绍了混沌时间序列的预测方法,并借助G-P算法、C-C方法和小数据量方法从武威盆地地下水位埋深一维时间序列中提取Lyapunov指数,计算结果表明此时间序列具有混沌特性,在此基础上建立了混沌时间序列的一阶加权局域预测模型,对武威盆地的地下水位埋深进行预测。结果表明,该模型预测精度较高,可有效的对地下水埋深的动态进行预报。Based on phase-space reconstruction theory and the predication theory of chaotic time series,this paper used the G-P algorithm,C-C methods and small data sets to extract Lyapunov index from one-dimensional time series of groundwater depth in Wuwei Basin,and the Lyapunov index indicated that the time series is possessed of chaotic state.Therefore,the first order weighted local predication model of chaotic time series was built,which was used to predicate the groundwater depth in Wuwei Basin and got the conclusion that the model had a high predication accuracy and could effectively predict the dynamic of groundwater depth.

关 键 词：混沌 时间序列 相空间重构 LYAPUNOV指数 局域法 

分 类 号：TV138[水利工程—水力学及河流动力学]