BP神经网络在西北地区集雨窖水水质评价中的应用

Rainwater Quality Comprehensive Assessment in the Northwestern Region by BP Networks

本研究以会宁地区集雨窖水水质中的CODMn、NH3-N、DO三个参评指标作为集雨窖水水质等级评价的输入特征值，建立了BP神经网络水质评价模型，得出集雨初期的窖水水质基本上为v类或超v类，随着时间的增长窖水水质有明显改善，可达到I～Ⅱ类水质。

The study was to establish BP neural network models, in which CODMn, NH3-N, DO in were used as the reference index for comprehensive assessment of water quality of harvested rainwater in northwest china, the harvested rainwater of huining area. The quality of initial harvested rainwater in water cellars were class V or much higher than water quality standard, but the water quality was ameliorate from inferior standard V to standard I or Ⅱ with time prolong.