基于人工神经网络的松花江朱顺屯断面的水质预测

Based on ANN of songhua river ZhuShunTun section water quality forecast

利用BP神经网络进行短期水质预报并进行水质预测后再评价对环境管理和规划有重要意义。本文建立了朱顺屯断面5个目标水质参数NH3-N、DO、高锰酸盐指数、TN和TP的BPANN水质预测模型,并对实例进行了验证,结果表明建立的推进式ANN可以应用于此类时间序列的环境预测,网络泛化性能好,能够满足实际应用。

Using BP neural network to carry on the short - term water quality forecast and predict water quality after the evaluation method of environment management and planning has important significance. This paper es- tablished the ZhuShunTun section five target water quality parameters NH3 - N, DO and CODmn, TN and TP BPANN of water quality prediction model, and validated, and the results show that the establishment of push type ANN can be applied to such time series prediction of environment, network generalization performance is good, can satisfy the practical application.