人工神经网络模型在太湖富营养化评价中的应用

根据湖泊富营养化程度影响因素多，评价因素与富营养化等级之间关系复杂且是非线性的特点，研制了一个能自动对湖泊富营养化程度做出正确评价的BP人工神经网络模型，并在太湖富营养化评价中得到了应用，结果表明：只要把观测数据提供给网络，借助计算机就可获得能客观地反映水质富营养化状况的评价结果；对于富营养化标准样本，一旦训练完毕，只需通过简单的加法和乘法运算，就可对湖泊水质富营养化程度进行评价．

Application of artificial neural network model to assessment of Taihu Lake eutrophication

The degree of lake eutrophication is affected by many factors. In consideration of the complicated non-linear characteristics of the relationships between the eutrophication degree and some related factors, a BP artificial neural network, which can give correct assessment of lake eutrophication automatically, is developed and applied to assessment of the eutrophication of the Taihu Lake. The result shows that the model can reflect the eutrophication degree of water quality of the Taihu Lake through the computer with observed data, and that, once the BP neural network is trained by use of the standard samples, the eutrophication assessment of the lake water quality can be made through simple operation of the network.