【摘要】 地下水水位动态特征综合反映了气象因子、水文因子、人为因子的干扰。应用偏最小二乘回归分析三江平原地下水动态变化,可有效地进行数据准备和输入因素的成分提取。提取出的成分具有线性无关的特点,对地下水动态变化有较好的解释能力,利于建模和预测。此方法另一特点是可以消除输入因素的多重共线性,不需要大量样本作为输入。算例表明,该方法用于地下水动态预测建模速度快、预测精度高,是一种行之有效的方法。更多还原

【Abstract】 Groundwater level dynamic state reflect meteorological factors, hydrologic factors, man-made factors’ disturbance synthetically. Using partial least squares regression to analyze groundwater level dynamic characteristics of Sanjiang Plain, This can effectively perform the data preparing and abstract components from input factors. The abstracted components, having the property of linearly independence, can explain the groundwater dynamic characteristics better and are more favorable for modeling and forecasting. Another feature of this method is that the multiple conjugate linear of input factors can be eliminated, so a lot of samples to input are not needed. The results of calculation example from a real groundwater system show that in short-term groundwater level forecasting the presented method is effectual, with this method the high modeling speed and high forecasting accuracy can be obtained.更多还原