基于模糊聚类的神经网络模型及其在渗流分析中的应用

本文采用模糊聚类理论方法对因子集进行模糊聚类 ,然后利用神经网络的方法建立样本因子集类别变量特征值与样本观测值之间的预测模型 ,提出了将模糊聚类、模糊模式识别以及神经网络三者有机结合的预测理论。并通过某大坝渗流计算实例对传统的统计预报模型和基于模糊聚类的神经网络预测模型进行了比较 ,结果表明后者的预报精度比前者要高。

Neural network model based on fuzzy clustering and its application in the analysis of seepage flow

The factor sets are fuzzy clustered by using fuzzy clustering theory and method, then the forecasting model between sample factor's class variable characteristic value and sample observation value is established by use of neural network method, and forecasting theory is presented together with fuzzy clustering, fuzzy pattern recognition and neural network. As an example, a comparison is made between traditional statistical model and neural network foreacastimg model based on fuzzy clustering, and the results show that the forecasting precision of the latter is higher than that of former.