Background: In the last two decades the using of artificial intelligence models for correct operation of the water treatment plant and maintain the stability of systems in favorable conditions, much researches has been done in this area. These models to simulate the behavior of water treatment systems can be used as an effective tool and to be used in the prediction of plant performance. The aim of this study was to evaluate of Khorramabad wastewater treatment plant performance (WWTP) using artificial intelligence network (AIN).
Materials and Methods: In this study, by using the AIN-LM and underlying the quality parameters measured at the entrance of plant (T, PH, DO, BOD, COD, TSS, TDS, NO3, PO4), the corresponding three parameters BOD, COD and TSS in the output of wastewater treatment plant was predicted. Statistical indicators used in this study were R, MSE and the software Matlab and SPSS (test T-test), respectively.
Results: Based on the results, BOD, COD and TSS, respectively, with a maximum R, 0/98, 0/91 and 0/92 for the train data and 0/5, 0/66 and 0/5 for the test data and minimum MSE, 3/5, 33/15 and 2/17 for the train data 11, 115 and 20/99 were predicted for the test data, and the results were acceptable. Also, by calculating the percent removal of pollutants in the output of plant was revealed that TSS had the maximum efficiency of pollutant removal in wastewater treatment plant and was equal to 87/68 %. Also, other amounts of pollutant were closed to TSS.
Conclusion: In this study, AIN-LM created a reliable tool for predicting the performance of Khorramabad wastewater treatment plant and could predict the quality of effluent on the basis of measured parameters. So, remove of pollutants through the results were obtained by using the AIN-LM network, showed that, it was a good model, so the observed data indicates that confirm of the performance this model, as well. Also, the reduction of qualitative values as standard values recommended by the DOE indicates that the relatively good performance of the WWTP.