Research on Prediction of Total Nitrogen Concentration in Wastewater Treatment Plant Effluent Based on CA-GRU
To accurately predict the total nitrogen concentration of effluent from a wastewater treatment plant,the publicly monitored effluent wastewater quality data from a wastewater treatment plant in Huma county is used as a sample for the study.A convolutional attention-gated recurrent unit(CA-GRU)network hybrid model is proposed.Firstly,using a time sliding window,the data is converted into a continuous feature map as input,from which abstract features are extracted.Then,these features are mapped into a network model.Finally,the gated recurrent unit(GRU)network model is used to obtain the predicted values.The experimental results show that the root mean square error(RMSE)of the CA-GRU model is 0.172 and the mean absolute percentage error(MAPE)is 0.010.This result is lower than the GRU network model by 0.108,0.016,lower than the convolutional neural network(CNN)-GRU model by 0.027,0.005,and lower than the Attention-GRU model by 0.065,0.007.This result shows that the CA-GRU model prediction effect is good,and the use of models such as CNN is conducive to reducing the interference of redundant information.The CA-GRU model can fully extract the characteristics of wastewater water quality data in time and space,and can more accurately predict the total nitrogen content of effluent water quality,which is of high value for application.
NETL
NSTL
万方数据
维普
