BP神经网络用于饮用水管网细菌总数预测

Prediction of HPC in Drinking Water Distribution Networks by BP Neural Network

为保障饮用水管网细菌学水质安全,并实现对细菌总数指标的预测,选定北方某市实验小区饮用水管网为研究对象,应用Matlab软件建立BP神经网络的细菌总数预测模型.结果表明,BP神经网络方法能较好地模拟复杂系统,模型精度较高;同时,建立BP神经网络模型,对水质指标间的相互作用进行模拟,拟合效果较好,并针对该实验管网给出了某些指标的限值。

To ensure the security of bacteria water quality in drinking water distribution network, and make prediction on the index of total bacteria counts, heterotrophic plate counts （ HPC ） prediction model of a distribution network of an experimental area in a northern city of China was constructed by Matlab software with the methods of back propagation（ BP ） neural network. The result shows that the method of BP neural network models complex system well with high precision. At the same time, BP neural network model was constructed to simulate the correlation of parameters and HPC. The effect of the simulation was good and the limit value of some parameters were provided.