基于黄金分割理论的改进BP网络地下水水质评价

Evaluation of Groundwater Quality based on Improved BP Artificial Neural Networks of the Golden Section Theory

利用VB语言编写附加动量的改进BP人工神经网络模型程序,并将其加载到Excel中,以湛江市区地下水为例研究水质状况。该模型采用黄金分割理论和试算相结合的方法对网络模型的隐含层节点数进行了优选,研究结果与其他方法相比显示:改进BP人工神经网络模型在地下水水质评价中能够很好地解决评价因子与水质等级间复杂的非线形关系,评价结果的精度有较大地提高。

In this paper,the Groundwater Quality status were studied,taking example for water in the urban of Zhanjiang and using VB language to design the improved BP Artificial Neural Networks model of additional momentum program,which was loaded into Excel.In this model,arithmetic based on golden section theory and trial value were used to optimize the amount of nodes in network＇s covert layer.the evaluation results were compared with other methods.It show that the evaluation of the improved BP Artificial Neural Networks which has additional momentum to groundwater quality can be able to resolve the complex non-linear relationship between the assessment factors and the water quality classifications,the accuracy of the evaluation results is improved significantly.