| 摘要: |
| 【目的】建立混沌时间序列的支持向量机预报模型，为地下水动态提供新的可行的预报方法。【方法】以重构相空间理论为基础，探讨了混沌时间序列的支持向量机预报模型的建模思路、特点及参数的选取，借助G-P算法、C-C方法和Wolf方法，计算了武威盆地3眼观测井地下水位埋深序列的Lyapunov指数，并利用自适应方法对支持向量机的参数进行了选择；基于高斯径向基核函数，建立了混沌时间序列的支持向量机预报模型。【结果】武威盆地地下水位埋深序列的 Lyapunov指数均大于0，表明该时间序列具有混沌特性；所建立的混沌-支持向量机模型可以用于武威盆地地下水位埋深预报，经过检验，武威盆地3眼观测井的预报精度分别为0.98，0.92和0.86，表明建立模型预报精度较为理想。【结论】建立了混沌-支持向量机模型，该模型可用于地下水位埋深动态预报。 |
| 关键词: 地下水动态 混沌 时间序列 相空间重构 Lyapunov指数 支持向量机 |
| DOI： |
| 分类号: |
| 基金项目: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) |
|
| Prediction model of chaotic based on support vector machine and its application to underground Water |
|
HU Guojie, WEI Xiaomei, CAI Mingke,et al
|
|
College of Water Resources and Architectural Engineering, Northwest A&F University
|
| Abstract: |
| 【Objective】The objective was to establish a chaotic based prediction model on support vector machine for the groundwater.【Method】Chaos theory and support vector machine have great capability of dealing with nonlinear matter.Based on the phase space reconstitution theory,the prediction model of chaos time series was built by using the support vector machine in this paper.The method,the characteristic,and the selecting of the key parameters in the modeling was discussed.With G-P arithmetic,C-C arithmetic and Wolf method,Lyapunov from 3 observation well of underground level depth in Wuwei basin was calculated.At the same time the method of adaptive support vector machine parameters was selected.A model was established about chaotic time series based on support vector machine with Gaussian radial basis function.【Result】Lyapunov indexes were greater than zero in Wuwei basin,indicating that the time series is chaotic.The model can be used on the Wuwei basin to predict the depth of groundwater level.After testing,the accuracy of three prediction results were 0.98,0.92 and 0.86,indicating that this mode has high prediction accuracy.【Conclusion】The prediction model of chaotic based on support vector machine can be used in forecasting the movement of underground water depth. |
| Key words: groundwater dynamic chaos time series phase space reconstruction Lyapunov exponent support vector machine |
