基于因子-神经网络预测南渡江海口段水质状况  

Water Quality Prediction in Haikou part of Nandujiang River Based on Factor Analysis and BP Neural Network

作　　者：王里奥[1] 任家宽[1] 刘阳生[2] 马培东[1] 

机构地区：[1]重庆大学资源及环境科学学院,重庆400044 [2]海南省环境科学研究院,海南海口570206

出　　处：《环境科学与管理》2008年第6期176-179,共4页Environmental Science and Management

摘　　要：针对海南省南渡江日益严重的水污染现状,根据南渡江海口段2001年-2006年的监测资料,应用因子分析法找出影响水质的主要污染指标,建立了基于BP神经网络的主要污染指标预测模型,并对其进行训练检验,结果证明该模型预测精度满足要求,再利用训练、检验后的模型对近几年主要污染因子进行预测,发现了变化规律,最后提出南渡江污染防治措施。Nandujiang River in Hainan Province is facing to the increasingly serious water pollution, according to the monitoring data for the Haikou part of Nandujiang River in 2001 -2006, the main indexes have been found by the Factor Analysis. the main indexes prediction model based on BP neural network has been established, and the result shows has higher precision. Moreover, the model is applied to predict the main pollution indexes for the Haikou part in the next years and find out its regularity. Countermeasures for water pollution prevention and control are put forward as well.

关 键 词：因子分析 BP神经网络 南渡江 水污染 

分 类 号：X522[环境科学与工程—环境工程]