Utilization of Neural Network to Predict Efficiency at the Shahid Rajayi Industerial Town Treatment Plant
Habib Pakrou1 * , Saeed Pakrou1 , Naser Mehrdadi2 and Mohammad Javad Amiri2
1
Civil Engineering – Environmental,
Aras International Campus,
Iran
2
Civil Engineering – Environmental,
University of Tehran,
Iran
http://dx.doi.org/10.12944/CWE.10.Special-Issue1.108
Predicting the efficiency of Shahid Rajayi industrial town treatment plant is performed in this study. The data are collected from the laboratory of the treatment plant. The correlation coefficient is performed for the candidate inputs and the treatment plant outputs in order to analyze the input and output of treatment plant and choosing the proper inputs. The input-output modeling is developed for each output COD, BOD and TSS using forward neural network. Five inputs of BOD, COD, TSS, pH and Temperature are used in this modeling. Levenberg–Marquardt algorithm is used to train the neural network. The comparison of neural networks with five inputs indicates a good correlation and it shows that we should use the minimum possible number of inputs in the structure of neural networks in the cases where the number of existing data is low for training the neural network.
Copy the following to cite this article:
Pakrou H, Pakrou S, Mehrdadi N, Amiri M. J Utilization of Neural Network to Predict Efficiency at the Shahid Rajayi Industerial Town Treatment Plant. Special Issue of Curr World Environ 2015;10(Special Issue May 2015). DOI:http://dx.doi.org/10.12944/CWE.10.Special-Issue1.108
Copy the following to cite this URL:
Pakrou H, Pakrou S, Mehrdadi N, Amiri M. J Utilization of Neural Network to Predict Efficiency at the Shahid Rajayi Industerial Town Treatment Plant. Special Issue of Curr World Environ 2015;10(Special Issue May 2015). Available from: http://cwejournal.org?p=700/
