The numerical simulation of wastewater treatment plant(WWTP)is becoming an important method to realize the intelligent management.However,the basic online data has vari-ous data quality problems such as missing and abnormal.In this paper,a method for cleaning the online data of WWTP based on LSTM is proposed.Firstly,preliminary anomaly detection was car-ried out based on statistical method.Then the anomaly is defined,and the anomaly detection results are corrected based on the logical rules in this field.Finally,LSTM model is trained to pre-dict the time of outliers(including missing values)to replace and complete cleaning.This cleaning method has been applied to the online system of several WWTP.The application results show that the method can effectively detect and remove abnormal data.Moreover,ensure the stability and reliability of the online simulation results.
维普
万方数据