摘要
人工神经网络是一门新兴的交叉学科,是处理非线性问题的有效方法。本文把影响地下水位的因素集作为网络的输入向量,地下水位本身作为网络的输出向量,构成了预测地下水位的BP网络模型。一个实例的应用实践表明,用BP网络预测地下水位较准确地反映了客观实际,比其它方法如回归模型具有较高的拟合精度和预测精度。
Being a developing cross-sctence,the artificial neural network(ANN)is an efficient method to deal with the nonlinear problems.In the paper,a BP network in which the factors to effect groundwater levelsare set as 1nput neurons and the groundwater level is set as output neuron,was established to forecast ground-water level.An application of the BP network in a real case indicates that the forecasting model with the BPnetwork may completely describe the relative between the groundwater level and 1ts affecting factors.It also indicates that the BP forecasting model is mor approximate to realities than other forecasting model such as linear or nonlinear regression models.
作者
刘国东
丁晶
Liu Guodong;Ding Jing(Dept.of Hydraulic Engineering,Sichuan Union University Chengdu 610065)
基金
国家自然科学基金
中国博士后科学基金资助项目
关键词
神经网络
BP网络
水位预测
回归模型
地下水
Artificial neural
networks.BP networks.Forecasting of grouodwater level
Regression model
作者简介
第一作者:刘国东,男,37岁,博士后研究人员、曾发表《BP网络在水女系统分析中的初步应用》等论文
