中文摘要：

分析了银北灌区地下水位较高的成因，研究了灌区地下水位年际和年内的动态变化规律，采用主成分分析方法，确定了灌区地下水位变化的主要影响因素，并用变速率的憎神经网络实现了对灌区的地下水位变化的模拟。结果表明灌区地下水位变化具有周期性、季节性的特点，并且可以被VLBP神经网络模型准确地模拟。

英文摘要：

Soil salinization of Yinbei irrigation region in Qingtongxia is very serious and restricts healthy development of agriculture. The main reason of soil salinization of irrigation region is excessively high groundwater table. In this paper, the reason of high groundwater table is analyzed, major factors influencing the groundwater table of irrigation region are ascertained, the change law of groundwater table within and between years is researched, and the groundwater table change is simulated with neural network model based on change velocity algorithm. These researches have laid theory foundation for controlling the groundwater table and soil salinization of Yinbei irrigation region.