污水预警中的化学需氧量(COD)预测技术  

The COD Forecasting Technology in Sewage Early Warning

作　　者：叶延亮[1] 庄严[2] 

机构地区：[1]北华大学科研处,吉林吉林132021 [2]北华大学电气信息工程学院,吉林吉林132021

出　　处：《化工自动化及仪表》2013年第12期1528-1531,1552,共5页Control and Instruments in Chemical Industry

基　　金：吉林省教育厅"十二五"规划项目(吉教科合字(2011)第127号)

摘　　要：针对污水处理厂监控系统对水质重要指标出水COD难以及时测量的问题,提出利用进水BOD及进水COD等辅助输入变量通过改进的Elman神经网络建立出水COD预测系统模型,采用某化工污水处理厂的实际运行数据作为训练和测试样本。仿真结果表明:改进结构的Elman神经网络模型对出水COD的预测准确率最高。Aiming at the monitoring system＇s difficulty in measuring COD in time in sewage treatment plant,applying the improved Elman neural network and the auxiliary input variables like BOD and COD in the inflow to establish COD forcasting model for the effulent was proposed; and meanwhile,the operating data from a chemical sewage treatment plant were employed as the training and test sample.The simulation results show that the improved Elman neural network model can predict the COD in effluent with the highest accuracy.

关 键 词：污水处理 COD 标准Elman神经网络 改进结构的Elman神经网络 LM算法 

分 类 号：TQ056.1[化学工程]