中文摘要：

为了预测地下水位动态变化，采用理论分析与实验方法，对多目标模糊识别的目标函数进行了全面分析，提出了不同权重情形下的模糊模式识别全解析模型，同时提出了模式样本模糊克隆的概念，并将提出的模型应用于地下水位动态预测。实验结果表明，三种预测模型测试样本的平均误差分别为4．53％、5．33％、4．61％。研究结果表明本文提出的地下水位动态预测方法是有效的。

英文摘要：

To forecast the groundwater level, the methods of theory analysis and experiment are employed. According to studying completely of multi-object fuzzy pattern recognition model＇s function, the complete analysis fuzzy recognition model is presented under different weight conditions, and the new conception of the pattern samples fuzzy colony is put out at the same time. The presented models in this paper are used to forecast level of groundwater. The experiment shows that the test sample＇s mean errors of three models are respectively 4.53%, 5.33%, 4.61%. The application example shows that the presented models in this paper are valid for predicting the groundwater level.