Prediction of groundwater depth based on generalized regression neural network in Jinghuiqu Irrigation District

摘要

Groundwater depth is the key parameter for groundwater resources management and exploitation. Because of the nonlinear relationship between groundwater depth and some parameters, Generalized Regression Neural Network (GRNN) was applied to predict groundwater depth in Jinghuiqu Irrigation District. The data were analyzed by SAS software system using principal component analysis (PCA) before building GRNN and the most influential parameters were selected. It is testified by instance that the relative error of GRNN which was built by selected parameters was smaller than that with raw data. Thus the prediction accuracy was enhanced by input data statistical analysis in advance.