基于BP神经网络的水质预测技术

Water Quality Prediction Technology Based on BP Neural Network

针对城市污水处理厂在生化处理污水过程中测量重要过程变量困难、无法及时检测的问题,提出了基于BP神经网络的预测技术.利用某化工污水处理厂的实际运行数据对提出的BP神经网络结构进行了训练和仿真.结果表明:通过合理选择辅助输入变量,可以实现对过程变量的在线预测.

According to the problem that it is very difficult to measure the important process variables and not in time to measure in the process of wastewater biochemical treatment,the paper proposed the prediction technology based on the BP neural network.Using actual data of a chemical sewage treatment plant,the proposed BP neural network structure was simulated and trained.The simulation results show that the online process variables prediction can be realized through the rational selection of auxiliary input variables.